1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 PP/01 Handle short writes when writing local log-files.
8 In practice, only affects FreeBSD (8 onwards).
9 Bugzilla 1053, with thanks to Dmitry Isaikin.
11 NM/01 Bugzilla 949 - Documentation tweak
13 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
16 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
18 PP/02 Implemented gsasl authenticator.
20 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
22 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
23 `pkg-config foo` for cflags/libs.
25 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
26 with rest of GSASL and with heimdal_gssapi.
28 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
29 `pkg-config foo` for cflags/libs for the TLS implementation.
31 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
32 properties get this fed in as external SSF. A number of robustness
33 and debugging improvements to the cyrus_sasl authenticator.
35 PP/08 cyrus_sasl server now expands the server_realm option.
37 PP/09 Bugzilla 1214 - Log authentication information in reject log.
38 Patch by Jeremy Harris.
40 PP/10 Added dbmjz lookup type.
42 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
44 PP/12 MAIL args handles TAB as well as SP, for better interop with
45 non-compliant senders.
46 Analysis and variant patch by Todd Lyons.
48 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
49 Bug report from Lars Müller <lars@samba.org> (via SUSE),
50 Patch from Dirk Mueller <dmueller@suse.com>
52 PP/13 tls_peerdn now print-escaped for spool files.
53 Observed some $tls_peerdn in wild which contained \n, which resulted
54 in spool file corruption.
56 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
57 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
58 or write after TLS renegotiation, which otherwise led to messages
61 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
62 as a tracking header (ie: a signed header comes before the signature).
63 Patch from Wolfgang Breyha.
65 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
66 comma-sep list; embedded commas doubled.
68 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
70 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
72 Report and patch from Dmitry Banschikov.
74 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
75 Removed SSL_clear() after SSL_new() which led to protocol negotiation
76 failures. We appear to now support TLS1.1+ with Exim.
78 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
79 lets Exim select keys and certificates based upon TLS SNI from client.
80 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
81 before an outbound SMTP session. New log_selector, +tls_sni.
83 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
84 NULL dereference. Report and patch from Alun Jones.
86 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
87 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
88 Not seeing resolver debug output on NetBSD, but suspect this is a
89 resolver implementation change.
91 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
92 Left warnings. Added "eximon gdb" invocation mode.
94 PP/21 Defaulting "accept_8bitmime" to true, not false.
96 PP/22 Added -bw for inetd wait mode support.
98 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
99 locate the relevant includes and libraries. Made this the default.
101 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
102 Bugzilla 1246, report and most of solution from Tomasz Kusy.
104 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
105 This may cause build issues on older platforms.
107 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
108 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
109 gnutls_require_mac & gnutls_require_protocols (no longer supported).
110 Added SNI support via GnuTLS too.
111 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
113 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
115 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
116 Applied second patch from Janne, implementing suggestion to default
117 multiple-strings-in-record handling to match SPF spec.
119 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
121 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
122 read-only, out of scope).
123 Patch from Wolfgang Breyha, report from Stuart Northfield.
125 PP/29 Fix three issues highlighted by clang analyser static analysis.
126 Only crash-plausible issue would require the Cambridge-specific
127 iplookup router and a misconfiguration.
128 Report from Marcin Mirosław.
130 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
132 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
133 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
134 As part of this, removing so much warning spew let me fix some minor
135 real issues in debug logging.
137 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
138 assignment on my part. Fixed.
140 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
141 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
142 Janne Snabb (who went above and beyond: thank you).
144 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
145 string otherwise requires a connection and a bunch more work and it's
146 relatively easy to get wrong. Should also expose TLS library linkage
149 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
150 64-bit ${eval} (JH/03).
156 PP/01 Solaris build fix for Oracle's LDAP libraries.
157 Bugzilla 1109, patch from Stephen Usher.
159 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
161 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
164 TF/02 Fix a couple more cases where we did not log the error message
165 when unlink() failed. See also change 4.74-TF/03.
167 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
168 lock up or crash if it happened to be inside a call to libc when it
169 got a SIGUSR1 from exiwhat.
171 The SIGUSR1 handler appends the current process status to the process
172 log which is later printed by exiwhat. It used to use the general
173 purpose logging code to do this, but several functions it calls are
174 not safe for signals.
176 The new output code in the SIGUSR1 handler is specific to the process
177 log, and simple enough that it's easy to inspect for signal safety.
178 Removing some special cases also simplifies the general logging code.
179 Removing the spurious timestamps from the process log simplifies
182 TF/04 Improved ratelimit ACL condition.
184 The /noupdate option has been deprecated in favour of /readonly which
185 has clearer semantics. The /leaky, /strict, and /readonly update modes
186 are mutually exclusive. The update mode is no longer included in the
187 database key; it just determines when the database is updated. (This
188 means that when you upgrde Exim will forget old rate measurements.)
190 Exim now checks that the per_* options are used with an update mode that
191 makes sense for the current ACL. For example, when Exim is processing a
192 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
193 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
194 must specify per_mail/readonly. If you omit the update mode it defaults to
195 /leaky where that makes sense (as before) or /readonly where required.
197 The /noupdate option is now undocumented but still supported for
198 backwards compatibility. It is equivalent to /readonly except that in
199 ACLs where /readonly is required you may specify /leaky/noupdate or
200 /strict/noupdate which are treated the same as /readonly.
202 A useful new feature is the /count= option. This is a generalization
203 of the per_byte option, so that you can measure the throughput of other
204 aggregate values. For example, the per_byte option is now equivalent
205 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
207 The per_rcpt option has been generalized using the /count= mechanism
208 (though it's more complicated than the per_byte equivalence). When it is
209 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
210 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
211 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
212 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
213 Note that using per_rcpt with a non-readonly update mode in more than
214 one ACL will cause the recipients to be double-counted. (The per_mail
215 and per_byte options don't have this problem.)
217 The handling of very low rates has changed slightly. If the computed rate
218 is less than the event's count (usually one) then this event is the first
219 after a long gap. In this case the rate is set to the same as this event's
220 count, so that the first message of a spam run is counted properly.
222 The major new feature is a mechanism for counting the rate of unique
223 events. The new per_addr option counts the number of different
224 recipients that someone has sent messages to in the last time period. It
225 behaves like per_rcpt if all the recipient addresses are different, but
226 duplicate recipient addresses do not increase the measured rate. Like
227 the /count= option this is a general mechanism, so the per_addr option
228 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
229 example, measure the rate that a client uses different sender addresses
230 with the options per_mail/unique=$sender_address. There are further
231 details in the main documentation.
233 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
235 TF/06 Removed a few PCRE remnants.
237 TF/07 Automatically extract Exim's version number from tags in the git
238 repository when doing development or release builds.
240 PP/02 Raise smtp_cmd_buffer_size to 16kB.
241 Bugzilla 879. Patch from Paul Fisher.
243 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
244 Heavily based on revision 40f9a89a from Simon Arlott's tree.
247 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
249 PP/05 Variable $av_failed, true if the AV scanner deferred.
250 Bugzilla 1078. Patch from John Horne.
252 PP/06 Stop make process more reliably on build failure.
253 Bugzilla 1087. Patch from Heiko Schlittermann.
255 PP/07 Make maildir_use_size_file an _expandable_ boolean.
256 Bugzilla 1089. Patch from Heiko Schlittermann.
258 PP/08 Handle ${run} returning more data than OS pipe buffer size.
259 Bugzilla 1131. Patch from Holger Weiß.
261 PP/09 Handle IPv6 addresses with SPF.
262 Bugzilla 860. Patch from Wolfgang Breyha.
264 PP/10 GnuTLS: support TLS 1.2 & 1.1.
266 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
269 PP/11 match_* no longer expand right-hand-side by default.
270 New compile-time build option, EXPAND_LISTMATCH_RHS.
271 New expansion conditions, "inlist", "inlisti".
273 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
275 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
277 PP/14 fix log_write() format string regression from TF/03.
278 Bugzilla 1152. Patch from Dmitry Isaikin.
284 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
286 PP/02 Harmonised TLS library version reporting; only show if debugging.
287 Layout now matches that introduced for other libraries in 4.74 PP/03.
289 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
291 PP/04 New "dns_use_edns0" global option.
293 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
296 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
297 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
299 TK/01 Updated PolarSSL code to 0.14.2.
300 Bugzilla 1097. Patch from Andreas Metzler.
302 PP/07 Catch divide-by-zero in ${eval:...}.
305 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
308 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
309 format-string attack -- SECURITY: remote arbitrary code execution.
311 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
312 time unintentionally subject to list matching rules, letting the header
313 cause arbitrary Exim lookups (of items which can occur in lists, *not*
314 arbitrary string expansion). This allowed for information disclosure.
316 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
317 INT_MIN/-1 -- value coerced to INT_MAX.
323 NM/01 Workround for PCRE version dependency in version reporting
326 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
327 This fixes portability to compilers other than gcc, notably
328 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
330 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
331 makefiles for portability to HP-UX and POSIX correctness.
333 PP/01 Permit LOOKUP_foo enabling on the make command-line.
334 Also via indented variable definition in the Makefile.
335 (Debugging by Oliver Heesakkers).
337 PP/02 Restore caching of spamd results with expanded spamd_address.
338 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
340 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
341 Improves build reliability. Fix from: Frank Elsner
343 NM/02 Fix wide character breakage in the rfc2047 coding
344 Fixes bug 1064. Patch from Andrey N. Oktyabrski
346 NM/03 Allow underscore in dnslist lookups
347 Fixes bug 1026. Patch from Graeme Fowler
349 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
350 Code patches from Adam Ciarcinski of NetBSD.
352 NM/04 Fixed exiqgrep to cope with mailq missing size issue
355 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
356 is logged, to avoid truncation. Patch from John Horne.
358 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
359 Patch from Jakob Hirsch.
361 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
362 SQL string expansion failure details.
363 Patch from Andrey Oktyabrski.
365 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
366 Patch from Simon Arlott.
368 PP/09 New lookups functionality failed to compile on old gcc which rejects
369 extern declarations in function scope.
370 Patch from Oliver Fleischmann
372 PP/10 Use sig_atomic_t for flags set from signal handlers.
373 Check getgroups() return and improve debugging.
374 Fixed developed for diagnosis in bug 927 (which turned out to be
377 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
378 Patch from Mark Zealey.
380 PP/12 Bugzilla 1056: Improved spamd server selection.
381 Patch from Mark Zealey.
383 PP/13 Bugzilla 1086: Deal with maildir quota file races.
384 Based on patch from Heiko Schlittermann.
386 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
387 Patch from Uwe Doering, sign-off by Michael Haardt.
389 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
390 variable declaration deep within a block. Bug and patch from
393 PP/15 lookups-Makefile IRIX compatibilty coercion.
395 PP/16 Make DISABLE_DKIM build knob functional.
397 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
398 Patch by Simon Arlott
400 TF/03 Fix valgrind.h portability to C89 compilers that do not support
401 variable argument macros. Our copy now differs from upstream.
407 TF/01 Failure to get a lock on a hints database can have serious
408 consequences so log it to the panic log.
410 TF/02 Log LMTP confirmation messages in the same way as SMTP,
411 controlled using the smtp_confirmation log selector.
413 TF/03 Include the error message when we fail to unlink a spool file.
415 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
416 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
417 for maintaining out-of-tree patches for some time.
419 PP/01 Bugzilla 139: Documentation and portability issues.
420 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
421 Handle per-OS dynamic-module compilation flags.
423 PP/02 Let /dev/null have normal permissions.
424 The 4.73 fixes were a little too stringent and complained about the
425 permissions on /dev/null. Exempt it from some checks.
426 Reported by Andreas M. Kirchwitz.
428 PP/03 Report version information for many libraries, including
429 Exim version information for dynamically loaded libraries. Created
430 version.h, now support a version extension string for distributors
431 who patch heavily. Dynamic module ABI change.
433 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
434 privilege escalation vulnerability whereby the Exim run-time user
435 can cause root to append content of the attacker's choosing to
438 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
441 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
442 If dropping privileges for untrusted macros, we disabled normal logging
443 on the basis that it would fail; for the Exim run-time user, this is not
444 the case, and it resulted in successful deliveries going unlogged.
445 Fixed. Reported by Andreas Metzler.
451 PP/01 Date: & Message-Id: revert to normally being appended to a message,
452 only prepend for the Resent-* case. Fixes regression introduced in
453 Exim 4.70 by NM/22 for Bugzilla 607.
455 PP/02 Include check_rfc2047_length in configure.default because we're seeing
456 increasing numbers of administrators be bitten by this.
458 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
460 PP/03 Bugzilla 994: added openssl_options main configuration option.
462 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
464 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
466 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
468 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
469 without a peer certificate, leading to a segfault because of an
470 assumption that peers always have certificates. Be a little more
471 paranoid. Problem reported by Martin Tscholak.
473 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
474 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
475 NB: ClamAV planning to remove STREAM in "middle of 2010".
476 CL also introduces -bmalware, various -d+acl logging additions and
477 more caution in buffer sizes.
479 PP/09 Implemented reverse_ip expansion operator.
481 PP/10 Bugzilla 937: provide a "debug" ACL control.
483 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
485 PP/12 Bugzilla 973: Implement --version.
487 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
489 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
491 PP/15 Bugzilla 816: support multiple condition rules on Routers.
493 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
494 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
495 ignore trailing whitespace.
497 JJ/02 prevent non-panic DKIM error from being sent to paniclog
499 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
502 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
503 Notification from Dr Andrew Aitchison.
505 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
506 ExtendedDetectionInfo response format.
507 Notification from John Horne.
509 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
512 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
513 XSL and documented dependency on system catalogs, with examples of how
516 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
519 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
520 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
521 configuration file which is writeable by the Exim user or group.
523 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
524 of configuration files to cover files specified with the -C option if
525 they are going to be used with root privileges, not just the default
528 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
529 option (effectively making it always true).
531 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
532 files to be used while preserving root privileges.
534 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
535 that rogue child processes cannot use them.
537 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
538 run-time user, instead of root.
540 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
541 Exim run-time user without dropping privileges.
543 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
544 result string, instead of calling string_vformat() twice with the same
547 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
548 for other users. Others should always drop root privileges if they use
549 -C on the command line, even for a whitelisted configure file.
551 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
553 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
559 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
560 $data_path, and $header_path variables; fixed documentation bugs and
563 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
564 exipick to access non-standard spools, including the "frozen" queue
567 NM/01 Bugzilla 965: Support mysql stored procedures.
568 Patch from Alain Williams
570 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
572 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
573 Patch from Andreas Metzler
575 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
576 Patch from Kirill Miazine
578 NM/05 Bugzilla 671: Added umask to procmail example.
580 JJ/03 installed exipick 20100323.0, fixing doc bug
582 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
583 directory. Notification and patch from Dan Rosenberg.
585 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
587 TK/02 Improve log output when DKIM signing operation fails.
589 MH/01 Treat the transport option dkim_domain as a colon separated
590 list, not as a single string, and sign the message with each element,
591 omitting multiple occurences of the same signer.
593 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
594 Bugzilla 985, 986. Patch by Simon Arlott
596 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
597 Patch by Simon Arlott
599 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
600 MBX locking. Notification from Dan Rosenberg.
606 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
608 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
610 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
612 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
614 NM/04 Bugzilla 915: Flag broken perl installation during build.
620 TK/01 Added patch by Johannes Berg that expands the main option
621 "spamd_address" if it starts with a dollar sign.
623 TK/02 Write list of recipients to X-Envelope-Sender header when building
624 the mbox-format spool file for content scanning (suggested by Jakob
627 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
628 (http://www.dcc-servers.net/) support via dccifd. Activated by
629 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
631 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
632 by Mark Daniel Reidel <mr@df.eu>.
634 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
635 When building exim an external PCRE library is now needed -
636 PCRE is a system library on the majority of modern systems.
637 See entry on PCRE_LIBS in EDITME file.
639 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
640 conversation. Added nologin parameter to request.
641 Patch contributed by Kirill Miazine.
643 TF/01 Do not log submission mode rewrites if they do not change the address.
645 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
647 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
648 log files in place. Contributed by Roberto Lima.
650 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
652 TF/03 Bugzilla 615: When checking the local_parts router precondition
653 after a local_part_suffix or local_part_prefix option, Exim now
654 does not use the address's named list lookup cache, since this
655 contains cached lookups for the whole local part.
657 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
658 Robert Millan. Documentation is in experimental-spec.txt.
660 TF/04 Bugzilla 668: Fix parallel build (make -j).
662 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
664 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
665 Patch provided by Jan Srzednicki.
667 TF/05 Leading white space used to be stripped from $spam_report which
668 wrecked the formatting. Now it is preserved.
670 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
671 that they are available at delivery time.
673 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
675 TF/08 TLS error reporting now respects the incoming_interface and
676 incoming_port log selectors.
678 TF/09 Produce a more useful error message if an SMTP transport's hosts
679 setting expands to an empty string.
681 NM/06 Bugzilla 744: EXPN did not work under TLS.
682 Patch provided by Phil Pennock.
684 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
685 Patch provided by Richard Godbee.
687 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
688 acl_smtp_notquit, added index entry.
690 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
691 Patch provided by Eugene Bujak.
693 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
694 accept(). Patch provided by Maxim Dounin.
696 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
697 Patch provided by Phil Pennock.
699 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
701 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
702 Patch provided by Brad "anomie" Jorsch.
704 NM/14 Bugzilla 622: Added timeout setting to transport filter.
705 Patch provided by Dean Brooks.
707 TK/05 Add native DKIM support (does not depend on external libraries).
709 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
710 Patch provided by Graeme Fowler.
712 NM/16 Bugzilla 851: Documentation example syntax fix.
714 NM/17 Changed NOTICE file to remove references to embedded PCRE.
716 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
719 NM/19 Bugzilla 745: TLS version reporting.
720 Patch provided by Phil Pennock.
722 NM/20 Bugzilla 167: bool: condition support.
723 Patch provided by Phil Pennock.
725 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
726 clients. Patch provided by Phil Pennock.
728 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
729 Patch provided by Brad "anomie" Jorsch.
731 NM/23 Bugzilla 687: Fix misparses in eximstats.
732 Patch provided by Heiko Schlittermann.
734 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
735 Patch provided by Heiko Schlittermann.
737 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
738 plus update to original patch.
740 NM/26 Bugzilla 799: Documentation correction for ratelimit.
742 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
743 Patch provided by David Brownlee.
745 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
747 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
749 NM/30 Bugzilla 888: TLS documentation bugfixes.
751 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
753 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
754 Unlike the original bugzilla I have changed all shell scripts in src tree.
756 NM/33 Bugzilla 898: Transport filter timeout fix.
757 Patch by Todd Rinaldo.
759 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mistmatches.
760 Patch by Serge Demonchaux.
762 NM/35 Bugzilla 39: Base64 decode bug fixes.
763 Patch by Jakob Hirsch.
765 NM/36 Bugzilla 909: Correct connect() call in dcc code.
767 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
769 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
771 NM/39 Bugzilla 911: Fixed MakeLinks build script.
777 TK/01 Add preliminary DKIM support. Currently requires a forked version of
778 ALT-N's libdkim that I have put here:
779 http://duncanthrax.net/exim-experimental/
781 Note to Michael Haardt: I had to rename some vars in sieve.c. They
782 were called 'true' and it seems that C99 defines that as a reserved
783 keyword to be used with 'bool' variable types. That means you could
784 not include C99-style headers which use bools without triggering
785 build errors in sieve.c.
787 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
788 as mailq or other aliases. Changed the --help handling significantly
789 to do whats expected. exim_usage() emits usage/help information.
791 SC/01 Added the -bylocaldomain option to eximstats.
793 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
795 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
797 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
803 PH/01 Another patch from the Sieve maintainer.
805 PH/02 When an IPv6 address is converted to a string for single-key lookup
806 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
807 dots are used instead of colons so that keys in lsearch files need not
808 contain colons. This was done some time before quoting was made available
809 in lsearch files. However, iplsearch files do require colons in IPv6 keys
810 (notated using the quote facility) so as to distinguish them from IPv4
811 keys. This meant that lookups for IP addresses in host lists did not work
812 for iplsearch lookups.
814 This has been fixed by arranging for IPv6 addresses to be expressed with
815 colons if the lookup type is iplsearch. This is not incompatible, because
816 previously such lookups could never work.
818 The situation is now rather anomolous, since one *can* have colons in
819 ordinary lsearch keys. However, making the change in all cases is
820 incompatible and would probably break a number of configurations.
822 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
825 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
826 conversion specification without a maximum field width, thereby enabling
827 a rogue spamd server to cause a buffer overflow. While nobody in their
828 right mind would setup Exim to query an untrusted spamd server, an
829 attacker that gains access to a server running spamd could potentially
830 exploit this vulnerability to run arbitrary code as the Exim user.
832 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
833 $primary_hostname instead of what libspf2 thinks the hosts name is.
835 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
836 a directory entry by the name of the lookup key. Previously, if a
837 symlink pointed to a non-existing file or a file in a directory that
838 Exim lacked permissions to read, a lookup for a key matching that
839 symlink would fail. Now it is enough that a matching directory entry
840 exists, symlink or not. (Bugzilla 503.)
842 PH/03 The body_linecount and body_zerocount variables are now exported in the
845 PH/04 Added the $dnslist_matched variable.
847 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
848 This means they are set thereafter only if the connection becomes
851 PH/06 Added the client_condition to authenticators so that some can be skipped
852 by clients under certain conditions.
854 PH/07 The error message for a badly-placed control=no_multiline_responses left
855 "_responses" off the end of the name.
857 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
859 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
860 (without spaces) instead of just copying the configuration text.
862 PH/10 Added the /noupdate option to the ratelimit ACL condition.
864 PH/11 Added $max_received_linelength.
866 PH/12 Added +ignore_defer and +include_defer to host lists.
868 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
869 way in which PCRE > 7.0 is built.
871 PH/14 Implemented queue_only_load_latch.
873 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
874 MAIL command. The effect was to mangle the value on 64-bit systems.
876 PH/16 Another patch from the Sieve maintainer.
878 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
880 PH/18 If a system quota error occurred while trying to create the file for
881 a maildir delivery, the message "Mailbox is full" was not appended to the
882 bounce if the delivery eventually timed out. Change 4.67/27 below applied
883 only to a quota excession during the actual writing of the file.
885 PH/19 It seems that peer DN values may contain newlines (and other non-printing
886 characters?) which causes problems in log lines. The DN values are now
887 passed through string_printing() before being added to log lines.
889 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
890 and InterBase are left for another time.)
892 PH/21 Added message_body_newlines option.
894 PH/22 Guard against possible overflow in moan_check_errorcopy().
896 PH/23 POSIX allows open() to be a macro; guard against that.
898 PH/24 If the recipient of an error message contained an @ in the local part
899 (suitably quoted, of course), incorrect values were put in $domain and
900 $local_part during the evaluation of errors_copy.
906 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
907 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
910 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
911 issue a MAIL command.
913 PH/02 In an ACL statement such as
915 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
917 if a client was not listed at all, or was listed with a value other than
918 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
919 the condition was not true (as it should be), so access was not denied.
920 The bug was that the ! inversion was incorrectly passed on to the second
921 item. This has been fixed.
923 PH/03 Added additional dnslists conditions == and =& which are different from
924 = and & when the dns lookup returns more than one IP address.
926 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
927 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
929 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
930 FSYNC, which compiles an option called disable_fsync that allows for
931 bypassing fsync(). The documentation is heavily laced with warnings.
933 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
935 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
936 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
937 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
938 including adding "make clean"; (3) Added -fPIC when compiling the test
939 dynamically loaded module, to get rid of a warning.
941 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
942 message fails, move_frozen_messages = true and ignore_bounce_errors_after
943 = 0s. The bug is otherwise harmless.
945 PH/07 There was a bug in the dovecot authenticator such that the value of
946 $auth1 could be overwritten, and so not correctly preserved, after a
947 successful authentication. This usually meant that the value preserved by
948 the server_setid option was incorrect.
950 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
952 PH/09 Installed PCRE release 7.0.
954 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
955 run for batched SMTP input. It is now run at the start of every message
956 in the batch. While fixing this I discovered that the process information
957 (output by running exiwhat) was not always getting set for -bs and -bS
958 input. This is fixed, and it now also says "batched" for BSMTP.
960 PH/11 Added control=no_pipelining.
962 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
963 patch, slightly modified), and move the expansion of helo_data till after
964 the connection is made in the smtp transport (so it can use these
967 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
969 PH/14 Added log_selector = +pid.
971 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
973 PH/16 Add ${if forany and ${if forall.
975 PH/17 Added dsn_from option to vary the From: line in DSNs.
977 PH/18 Flush SMTP output before performing a callout, unless control =
978 no_callout_flush is set.
980 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
981 was true (the default) a successful delivery failed to delete the retry
982 item, thus causing premature timeout of the address. The bug is now
985 PH/20 Added hosts_avoid_pipelining to the smtp transport.
987 PH/21 Long custom messages for fakedefer and fakereject are now split up
988 into multiline reponses in the same way that messages for "deny" and
989 other ACL rejections are.
991 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
992 with slight modification.
994 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
995 draft, changing the syntax and factoring some duplicate code".
997 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
998 for deliveries of the second and subsequent messages over the same SMTP
1001 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1002 ${reduce, with only minor "tidies".
1004 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1006 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1007 expansion side effects.
1009 PH/27 When a message times out after an over-quota error from an Exim-imposed
1010 quota, the bounce message says "mailbox is full". This message was not
1011 being given when it was a system quota that was exceeded. It now should
1014 MH/03 Made $recipients available in local_scan(). local_scan() already has
1015 better access to the recipient list through recipients_list[], but
1016 $recipients can be useful in postmaster-provided expansion strings.
1018 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1019 in the case of a MAIL command with additional options following the
1020 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1021 were accidentally chopped off.
1023 PH/29 SMTP synchronization checks are implemented when a command is read -
1024 there is a check that no more input is waiting when there shouldn't be
1025 any. However, for some commands, a delay in an ACL can mean that it is
1026 some time before the response is written. In this time, more input might
1027 arrive, invalidly. So now there are extra checks after an ACL has run for
1028 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1029 pipelining has not been advertised.
1031 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1033 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1034 returned in the SMTP response when smtp_return_error_details was false.
1035 This has been fixed.
1037 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1038 instead of the C I/O that was originally supplied, because problems were
1039 reported on Solaris.
1041 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1042 Exim which did not show up earlier: it was assuming that a call to
1043 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1044 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1045 was a macro that became an assignment, so it seemed to work. This has
1046 changed to a proper function call with a void return, hence the compile
1047 error. Exim's code has been fixed.
1049 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1052 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1054 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1056 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1057 because it turns out to be too big or there is a timeout) there is no
1058 "Completed" line in the log. When some messages of this type were
1059 selected by exigrep, they were listed as "not completed". Others were
1060 picked up by some special patterns. I have improved the selection
1061 criteria to be more general.
1063 PH/38 The host_find_failed option in the manualroute router can now be set
1064 to "ignore", to completely ignore a host whose IP address cannot be
1065 found. If all hosts are ignored, the behaviour is controlled by the new
1066 host_all_ignored option.
1068 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1069 homing or because of multiple MX records with /mx) generated more than
1070 one IP address, and the following item turned out to be the local host,
1071 all the secondary addresses of the first item were incorrectly removed
1072 from the list, along with the local host and any following hosts (which
1073 is what is supposed to happen).
1075 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1076 whoever called Exim into the -H file. In the case of the daemon it was
1077 behaving confusingly. When first started, it used values for whoever
1078 started the daemon, but after a SIGHUP it used the Exim user (because it
1079 calls itself on a restart). I have changed the code so that it now always
1082 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1083 message are rejected with the same error (e.g. no authentication or bad
1084 sender address), and a DATA command is nevertheless sent (as can happen
1085 with PIPELINING or a stupid MUA), the error message that was given to the
1086 RCPT commands is included in the rejection of the DATA command. This is
1087 intended to be helpful for MUAs that show only the final error to their
1090 PH/42 Another patch from the Sieve maintainer.
1092 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1093 Eximstats - Fixed some broken HTML links and added missing column headers
1095 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1096 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1098 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1100 PH/43 Yet another patch from the Sieve maintainer.
1102 PH/44 I found a way to check for a TCP/IP connection going away before sending
1103 the response to the final '.' that terminates a message, but only in the
1104 case where the client has not sent further data following the '.'
1105 (unfortunately, this is allowed). However, in many cases there won't be
1106 any further data because there won't be any more messages to send. A call
1107 to select() can be used: if it shows that the input is "ready", there is
1108 either input waiting, or the socket has been closed. An attempt to read
1109 the next input character can distinguish the two cases. Previously, Exim
1110 would have sent an OK response which the client would never have see.
1111 This could lead to message repetition. This fix should cure that, at
1112 least in a lot of common cases.
1114 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1115 advertised in response to EHLO.
1121 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1122 fixed by 4.65/MH/01 (is this a record?) are fixed:
1124 (i) An empty string was always treated as zero by the numeric comparison
1125 operators. This behaviour has been restored.
1127 (ii) It is documented that the numeric comparison operators always treat
1128 their arguments as decimal numbers. This was broken in that numbers
1129 starting with 0 were being interpreted as octal.
1131 While fixing these problems I realized that there was another issue that
1132 hadn't been noticed. Values of message_size_limit (both the global option
1133 and the transport option) were treated as octal if they started with 0.
1134 The documentation was vague. These values are now always treated as
1135 decimal, and I will make that clear in the documentation.
1141 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1142 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1145 MH/01 Don't check that the operands of numeric comparison operators are
1146 integers when their expansion is in "skipping" mode (fixes bug
1147 introduced by 4.64-PH/07).
1149 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1150 child addresses, Exim now panics and dies. Previously, because the count
1151 is held in a short int, deliveries were likely to be lost. As such a
1152 large number of recipients for a single message is ridiculous
1153 (performance will be very, very poor), I have chosen to impose a limit
1154 rather than extend the field.
1160 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1161 leftover -K file (the existence of which was triggered by #402).
1162 While we were at it, introduced process PID as part of the -K
1163 filename. This should rule out race conditions when creating
1166 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1167 processing considerably. Previous code took too long for large mails,
1168 triggering a timeout which in turn triggers #401.
1170 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1171 in the DK code in transports.c. sendfile() is not really portable,
1172 hence the _LINUX specificness.
1174 TF/01 In the add_headers option to the mail command in an Exim filter,
1175 there was a bug that Exim would claim a syntax error in any
1176 header after the first one which had an odd number of characters
1179 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1180 callout verification, Exim cached a "reject" for the entire domain. This
1181 is correct for most verifications, but it is not correct for a recipient
1182 verification with use_sender or use_postmaster set, because in that case
1183 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1184 case of MAIL FROM:<> rejection from other early rejections (e.g.
1185 rejection of HELO). When verifying a recipient using a non-null MAIL
1186 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1187 Whatever the result of the callout, the value of the domain cache is
1188 left unchanged (for any other kind of callout, getting as far as trying
1189 RCPT means that the domain itself is ok).
1191 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1194 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1195 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1196 session, the daemon ignores ECONNECT errors and logs others; it now
1197 ignores EPIPE as well.
1199 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1200 (quoted-printable decoding).
1202 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1203 later the small subsequent patch to fix an introduced bug.
1205 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1207 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1209 PH/08 An error is now given if message_size_limit is specified negative.
1211 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1212 to be given (somewhat) arbitrary names.
1214 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1217 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1218 miscellaneous code fixes
1220 PH/10 Added the log_reject_target ACL modifier to specify where to log
1223 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1224 hostname. This is wrong, because it relates to the incoming message (and
1225 probably the interface on which it is arriving) and not to the outgoing
1226 callout (which could be using a different interface). This has been
1227 changed to use the value of the helo_data option from the smtp transport
1228 instead - this is what is used when a message is actually being sent. If
1229 there is no remote transport (possible with a router that sets up host
1230 addresses), $smtp_active_hostname is used.
1232 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1233 tweaks were necessary in order to get it to work (see also 21 below):
1234 (a) The code assumed that strncpy() returns a negative number on buffer
1235 overflow, which isn't the case. Replaced with Exim's string_format()
1237 (b) There were several signed/unsigned issues. I just did the minimum
1238 hacking in of casts. There is scope for a larger refactoring.
1239 (c) The code used strcasecmp() which is not a standard C function.
1240 Replaced with Exim's strcmpic() function.
1241 (d) The code set only $1; it now sets $auth1 as well.
1242 (e) A simple test gave the error "authentication client didn't specify
1243 service in request". It would seem that Dovecot has changed its
1244 interface. Fortunately there's a specification; I followed it and
1245 changed what the client sends and it appears to be working now.
1247 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1250 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1251 address A is aliased to B and C, where B exists and C does not. Without
1252 -v the output is "A verified" because verification stops after a
1253 successful redirection if more than one address is generated. However,
1254 with -v the child addresses are also verified. Exim was outputting "A
1255 failed to verify" and then showing the successful verification for C,
1256 with its parentage. It now outputs "B failed to verify", showing B's
1257 parentage before showing the successful verification of C.
1259 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1260 look up a TXT record in a specific list after matching in a combined
1263 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1264 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1265 they consult the DNS. I had assumed they would set it the way they
1266 wanted; and indeed my experiments on Linux seem to show that in some
1267 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1268 To be on the safe side, however, I have now made the interface to
1269 host_find_byname() similar to host_find_bydns(), with an argument
1270 containing the DNS resolver options. The host_find_byname() function now
1271 sets these options at its start, just as host_find_bydns() does. The smtp
1272 transport options dns_qualify_single and dns_search_parents are passed to
1273 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1274 of host_find_byname() use the default settings of RES_DEFNAMES
1275 (qualify_single) but not RES_DNSRCH (search_parents).
1277 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1278 spool_read_header() do less string testing, by means of a preliminary
1279 switch on the second character of optional "-foo" lines. (This is
1280 overdue, caused by the large number of possibilities that now exist.
1281 Originally there were few.) While I was there, I also converted the
1282 str(n)cmp tests so they don't re-test the leading "-" and the first
1283 character, in the hope this might squeeze out yet more improvement.
1285 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1286 flag allowing group syntax was set by the header_syntax check but not
1287 turned off, possible causing trouble later; (2) The flag was not being
1288 set at all for the header_verify test, causing "group"-style headers to
1289 be rejected. I have now set it in this case, and also caused header_
1290 verify to ignore an empty address taken from a group. While doing this, I
1291 came across some other cases where the code for allowing group syntax
1292 while scanning a header line wasn't quite right (mostly, not resetting
1293 the flag correctly in the right place). These bugs could have caused
1294 trouble for malformed header lines. I hope it is now all correct.
1296 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1297 with the "reply" argument non-NULL. The code, however (which originally
1298 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1299 but it didn't always do it. This confused somebody who was copying the
1300 code for some other use. I have removed all the tests.
1302 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1303 feature that was used to support insecure browsers during the U.S. crypto
1304 embargo. It requires special client support, and Exim is probably the
1305 only MTA that supported it -- and would never use it because real RSA is
1306 always available. This code has been removed, because it had the bad
1307 effect of slowing Exim down by computing (never used) parameters for the
1308 RSA_EXPORT functionality.
1310 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1311 authenticator to fail if there's a tab character in the incoming data
1312 (there should never be unless someone is messing about, as it's supposed
1313 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1314 if the connection is using TLS or if the remote IP is the same as the
1315 local IP, and the "valid-client-cert option" if a client certificate has
1318 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1319 authenticators. This can be used for authorization after authentication
1320 succeeds. (In the case of plaintext, it servers for both authentication
1323 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1324 if any retry times were supplied.
1326 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1327 connection, where there is no client IP address to check. In this
1328 situation, the verify now always succeeds.
1330 PH/25 Applied John Jetmore's -Mset patch.
1332 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1334 PH/27 In a string expansion for a processed (not raw) header when multiple
1335 headers of the same name were present, leading whitespace was being
1336 removed from all of them, but trailing whitespace was being removed only
1337 from the last one. Now trailing whitespace is removed from each header
1338 before concatenation. Completely empty headers in a concatenation (as
1339 before) are ignored.
1341 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1342 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1344 PH/29 [Removed. This was a change that I later backed out, and forgot to
1345 correct the ChangeLog entry (that I had efficiently created) before
1346 committing the later change.]
1348 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1349 address errors (4xx response to RCPT) over the same connection as other
1350 messages routed to the same hosts. Such deliveries are always "forced",
1351 so retry times are not inspected. This resulted in far too many retries
1352 for the affected addresses. The effect occurred only when there were more
1353 hosts than the hosts_max_try setting in the smtp transport when it had
1354 the 4xx errors. Those hosts that it had tried were not added to the list
1355 of hosts for which the message was waiting, so if all were tried, there
1356 was no problem. Two fixes have been applied:
1358 (i) If there are any address or message errors in an SMTP delivery, none
1359 of the hosts (tried or untried) are now added to the list of hosts
1360 for which the message is waiting, so the message should not be a
1361 candidate for sending over the same connection that was used for a
1362 successful delivery of some other message. This seems entirely
1363 reasonable: after all the message is NOT "waiting for some host".
1364 This is so "obvious" that I'm not sure why it wasn't done
1365 previously. Hope I haven't missed anything, but it can't do any
1366 harm, as the worst effect is to miss an optimization.
1368 (ii) If, despite (i), such a delivery is accidentally attempted, the
1369 routing retry time is respected, so at least it doesn't keep
1370 hammering the server.
1372 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1373 in ${readsocket because some servers need this prod.
1375 PH/32 Added some extra debug output when updating a wait-xxx database.
1377 PH/33 The hint "could be header name not terminated by colon", which has been
1378 given for certain expansion errors for a long time, was not being given
1379 for the ${if def:h_colon_omitted{... case.
1381 PH/34 The spec says: "With one important exception, whenever a domain list is
1382 being scanned, $domain contains the subject domain." There was at least
1383 one case where this was not true.
1385 PH/35 The error "getsockname() failed: connection reset by peer" was being
1386 written to the panic log as well as the main log, but it isn't really
1387 panic-worthy as it just means the connection died rather early on. I have
1388 removed the panic log writing for the ECONNRESET error when getsockname()
1391 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1392 runs only) independently of the message's sender address. This meant
1393 that, if the 4xx error was in fact related to the sender, a different
1394 message to the same recipient with a different sender could confuse
1395 things. In particualar, this can happen when sending to a greylisting
1396 server, but other circumstances could also provoke similar problems.
1397 I have changed the default so that the retry time for these errors is now
1398 based a combination of the sender and recipient addresses. This change
1399 can be overridden by setting address_retry_include_sender=false in the
1402 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1403 remote server are returned as part of bounce messages. This was not
1404 happening for LMTP over a pipe (the lmtp transport), but now it is the
1405 same for both kinds of LMTP.
1407 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1408 in header lines that were in fact CNAMEs. This is no longer the case.
1410 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1411 and queue runs started by the daemon processed all messages. This has
1412 been fixed so that -R and -S can now usefully be given with -q<time>.
1414 PH/40 Import PCRE release 6.7 (fixes some bugs).
1416 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1418 PH/42 Give an error if -q is specified more than once.
1420 PH/43 Renamed the variables $interface_address and $interface_port as
1421 $received_ip_address and $received_port, to make it clear that these
1422 values apply to message reception, and not to the outgoing interface when
1423 a message is delivered. (The old names remain recognized, of course.)
1425 PH/44 There was no timeout on the connect() call when using a Unix domain
1426 socket in the ${readsocket expansion. There now is.
1428 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1429 be meaningful with "accept".
1431 SC/01 Eximstats V1.43
1432 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1434 SC/02 Eximstats V1.44
1435 Use a glob alias rather than an array ref in the generated
1436 parser. This improves both readability and performance.
1438 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1439 Collect SpamAssassin and rejection statistics.
1440 Don't display local sender or destination tables unless
1441 there is data to show.
1442 Added average volumes into the top table text output.
1444 SC/04 Eximstats V1.46
1445 Collect data on the number of addresses (recipients)
1446 as well as the number of messages.
1448 SC/05 Eximstats V1.47
1449 Added 'Message too big' to the list of mail rejection
1450 reasons (thanks to Marco Gaiarin).
1452 SC/06 Eximstats V1.48
1453 Mainlog lines which have GMT offsets and are too short to
1454 have a flag are now skipped.
1456 SC/07 Eximstats V1.49 (Alain Williams)
1457 Added the -emptyok flag.
1459 SC/08 Eximstats V1.50
1460 Fixes for obtaining the IP address from reject messages.
1462 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1463 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1464 whitesspace changes from 4.64-PH/27
1466 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1469 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1470 are found, allow negative numbers in numeric criteria)
1472 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1474 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1475 to match changes made in 4.64-PH/43
1477 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1479 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1480 "domain" is actually an IP address, and force a failure. This locks out
1481 those revolvers/nameservers that support "A-for-A" lookups, in
1482 contravention of the specifications.
1484 PH/48 When a host name was looked up from an IP address, and the subsequent
1485 forward lookup of the name timed out, the host name was left in
1486 $sender_host_name, contrary to the specification.
1488 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1489 restricted to single-key lookups, Exim was not diagnosing an error if
1490 * or *@ was used with a query-style lookup.
1492 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1494 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1495 long ago, but noone interested enough thought of it. Let's just say that
1496 the "1.1" means that there are some new functions that weren't there at
1497 some point in the past.
1499 PH/51 Error processing for expansion failure of helo_data from an smtp
1500 transport during callout processing was broken.
1502 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1503 tested/used via the -bh/-bhc/-bs options.
1505 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1506 bug, fixed in subsequent PCRE releases).
1508 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1509 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1511 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1517 SC/01 Use a glob alias rather than an array ref in eximstats generated
1518 parser. This improves both readability and performance.
1520 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1521 Don't display local sender or destination tables in eximstats unless
1522 there is data to show.
1523 Added average volumes into the eximstats top table text output.
1525 SC/03 Collect data on the number of addresses (recipients) as well
1526 as the number of messages in eximstats.
1528 TF/01 Correct an error in the documentation for the redirect router. Exim
1529 does (usually) call initgroups() when daemonizing.
1531 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1532 with consistent privilege compared to when running as a daemon.
1534 TF/03 Note in the spec that $authenticated_id is not set for local
1535 submissions from trusted users.
1537 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1538 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1540 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1541 by adding some example configuration directives to the default
1542 configuration file. A little bit of work is required to uncomment the
1543 directives and define how usernames and passwords are checked, but
1544 there is now a framework to start from.
1546 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1547 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1548 without this. I don't know how relevant this is to other LDAP libraries.
1550 PH/02 Add the verb name to the "unknown ACL verb" error.
1552 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1554 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1556 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1557 directory not expanded when it should be if an expanded home directory
1558 was set for the address (which is overridden by the transport).
1560 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1563 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1564 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1565 because it is too late at that time, and has no effect.
1567 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1568 security issue with \' (bugzilla #107). I could not use the
1569 PQescapeStringConn() function, because it needs a PGconn value as one of
1572 PH/08 When testing addresses using -bt, indicate those final addresses that
1573 are duplicates that would not cause an additional delivery. At least one
1574 person was confused, thinking that -bt output corresponded to deliveries.
1575 (Suppressing duplicates isn't a good idea as you lose the information
1576 about possibly different redirections that led to the duplicates.)
1578 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1579 systems where poll() doesn't work, in particular OS X.
1581 PH/10 Added more information to debugging output for retry time not reached.
1583 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1584 operations in malware.c.
1586 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1589 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1590 syslog_duplication set false, log lines that would normally be written
1591 both the the main log and to the reject log were not written to syslog at
1594 PH/14 In the default configuration, change the use of "message" in ACL warn
1595 statements to "add_header".
1597 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1598 not followed by a command (e.g. "seen endif").
1600 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1601 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1604 PH/17 Added extra conditions to the default value of delay_warning_condition
1608 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1609 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1610 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1613 The Auto-Submitted: and various List- headers are standardised, whereas I
1614 don't think Precedence: ever was.
1616 PH/18 Refactored debugging code in route_finduser() to show more information,
1617 in particular, the error code if getpwnam() issues one.
1619 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1620 This is apparently needed in addition to the PH/07 change above to avoid
1621 any possible encoding problems.
1623 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1624 but not after initializing Perl.
1626 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1627 output them only if debugging. By default they are written stderr,
1628 apparently, which is not desirable.
1630 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1633 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1636 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1638 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1639 authenticated or an ident call has been made. Suppress the default
1640 values for $authenticated_id and $authenticated_sender (but permit -oMai
1641 and -oMas) when testing with -bh.
1643 PH/24 Re-jigged the order of the tests in the default configuration so that the
1644 tests for valid domains and recipients precede the DNS black list and CSA
1645 tests, on the grounds that those ones are more expensive.
1647 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1648 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1649 This bug exists in every version of Exim that I still have, right back to
1652 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1653 However, an attempt to turn on case-sensitivity in a regex key by
1654 including (?-i) didn't work because the subject string was already
1655 lowercased, and the effects were non-intuitive. It turns out that a
1656 one-line patch can be used to allow (?-i) to work as expected.
1662 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1663 other effects) broke the use of negated acl sub-conditions.
1665 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1668 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1669 "Deny" causes Exim to reject the incoming connection with a 554 error.
1670 Unfortunately, if there is a major crisis, such as a disk failure,
1671 tcp-wrappers gives "deny", whereas what one would like would be some
1672 kind of temporary error. A kludge has been added to help with this.
1673 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1674 554 error is used if errno is still zero or contains ENOENT (which occurs
1675 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1678 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1680 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1681 errors. Otherwise a message that provokes a temporary error (when other
1682 messages do not) can cause a whole host to time out.
1684 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1685 addresses were routed directly to files or pipes from a redirect router.
1686 File deliveries just didn't batch; pipe deliveries might have suffered
1689 PH/06 A failure to get a lock for a hints database would erroneously always say
1690 "Failed to get write lock", even when it was really a read lock.
1692 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1693 of 0600. This has been changed to use the value of the lockfile_mode
1694 option (which defaults to 0600).
1696 PH/08 Applied small patch from the Sieve maintainer.
1698 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1699 folder from quota calculations, a direct delivery into this folder messed
1700 up the contents of the maildirsize file. This was because the regex was
1701 used only to exclude .Trash (or whatever) when the size of the mailbox
1702 was calculated. There was no check that a delivery was happening into an
1703 excluded directory. This bug has been fixed by ignoring all quota
1704 processing for deliveries into excluded directories.
1706 PH/10 Added the maildirfolder_create_regex option to appendfile.
1712 PH/01 The code for finding all the local interface addresses on a FreeBSD
1713 system running IPv6 was broken. This may well have applied to all BSD
1714 systems, as well as to others that have similar system calls. The broken
1715 code found IPv4 interfaces correctly, but gave incorrect values for the
1716 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1717 that it would not match correctly against @[] and not recognize the IPv6
1720 PH/02 The ipliteral router was not recognizing addresses of the form user@
1721 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1723 PH/03 Added disable_ipv6.
1725 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1726 decoded form, because it is most often used to construct To: headers
1727 lines in autoreplies, and the decoded form may well be syntactically
1728 invalid. However, $reply_address has leading white space removed, and all
1729 newlines turned into spaces so that the autoreply transport does not
1732 PH/05 If group was specified without a user on a router, and no group or user
1733 was specified on a transport, the group from the router was ignored.
1735 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1736 for visible compile-time settings that can be used to change these
1737 numbers, for those that want even more. Backwards compatibility with old
1738 spool files has been maintained. However, going back to a previous Exim
1739 release will lost any variables that are in spool files.
1741 PH/07 Two small changes when running in the test harness: increase delay when
1742 passing a TCP/IP connection to a new process, in case the original
1743 process has to generate a bounce, and remove special handling of
1744 127.0.0.2 (sic), which is no longer necessary.
1746 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1747 be the same on different OS.
1749 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1752 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1753 whether --show-vars was specified or not
1755 JJ/02 exipick: Added support for new ACL variable spool format introduced
1758 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1759 syntactically invalid From: or Reply-to: line, and a filter used this to
1760 generate an autoreply, and therefore failed to obtain an address for the
1761 autoreply, Exim could try to deliver to a non-existent relative file
1762 name, causing unrelated and misleading errors. What now happens is that
1763 it logs this as a hard delivery error, but does not attempt to create a
1766 PH/11 The exinext utility has a -C option for testing purposes, but although
1767 the given file was scanned by exinext itself; it wasn't being passed on
1768 when Exim was called.
1770 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1771 an end-of-file indication when reading a command response.
1773 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1774 compiled. In many other places in Exim, IPv6 addresses are always
1775 recognized, so I have changed this. It also means that IPv4 domain
1776 literals of the form [IPV4:n.n.n.n] are now always recognized.
1778 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1779 used if the router is not running as root, for example, when verifying at
1780 ACL time, or when using -bh. The debugging output from this situation was
1781 non-existent - all you got was a failure to exec. I have made two
1784 (a) Failures to set uid/gid, the current directory, or a process leader
1785 in a subprocess such as that created by queryprogram now generate
1786 suitable debugging ouput when -d is set.
1788 (b) The queryprogram router detects when it is not running as root,
1789 outputs suitable debugging information if -d is set, and then runs
1790 the subprocess without attempting to change uid/gid.
1792 PH/15 Minor change to Makefile for building test_host (undocumented testing
1795 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1796 additional section of a DNS packet that returns MX or SRV records.
1797 Instead, it always explicitly searches for A/AAAA records. This avoids
1798 major problems that occur when a DNS server includes only records of one
1799 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1800 fixed another bug: if SRV records were looked up and the corresponding
1801 address records were *not* found in the additional section, the port
1802 values from the SRV records were lost.
1804 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1805 using the correct key (the original address) when searching the retry
1806 rules in order to find which one to use for generating the retry hint.
1808 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1809 adding the default one. Similarly, if it contains a Reply-To: header, the
1810 errors_reply_to option, if set, is not used.
1812 PH/19 When calculating a retry time, Exim used to measure the "time since
1813 failure" by looking at the "first failed" field in the retry record. Now
1814 it does not use this if it is later than than the arrival time of the
1815 message. Instead it uses the arrival time. This makes for better
1816 behaviour in cases where some deliveries succeed, thus re-setting the
1817 "first failed" field. An example is a quota failure for a huge message
1818 when small messages continue to be delivered. Without this change, the
1819 "time since failure" will always be short, possible causing more frequent
1820 delivery attempts for the huge message than are intended.
1821 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1823 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1824 $1, $2, $3) because the numerical variables can be reset during some
1825 expansion items (e.g. "match"), thereby losing the authentication data.
1827 PH/21 Make -bV show the size of off_t variables so that the test suite can
1828 decide whether to run tests for quotas > 2G.
1830 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1831 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1832 filecount value is greater than 2G or if a quota value is greater than 2G
1833 on a system where the size of off_t is not greater than 4, a panic error
1836 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1837 never match. The debug and -bh output now contains an explicit error
1838 message indicating a malformed IPv4 address or mask.
1840 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1841 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1842 PH/23 above applies.
1844 PH/25 Do not write to syslog when running in the test harness. The only
1845 occasion when this arises is a failure to open the main or panic logs
1846 (for which there is an explicit test).
1848 PH/26 Added the /no_tell option to "control=freeze".
1850 PH/27 If a host name lookup failed very early in a connection, for example, if
1851 the IP address matched host_lookup and the reverse lookup yielded a name
1852 that did not have a forward lookup, an error message of the form "no IP
1853 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1854 could be logged. Now it outputs the IP address instead of "NULL".
1856 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1857 allows the sender and the authenticated sender to be set when
1858 submitting a message from within Exim. Since child_open_exim() is
1859 documented for local_scan(), the new function should be too.
1861 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1862 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1863 results in an empty string is now treated as unset.
1865 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1867 PH/31 Added sender_verify_fail logging option.
1869 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1870 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1871 tidied the source and removed it altogether.
1873 PH/33 When a queue run was abandoned because the load average was too high, a
1874 log line was always written; now it is written only if the queue_run log
1875 selector is set. In addition, the log line for abandonment now contains
1876 information about the queue run such as the pid. This is always present
1877 in "start" and "stop" lines but was omitted from the "abandon" line.
1879 PH/34 Omit spaces between a header name and the colon in the error message that
1880 is given when verify = headers_syntax fails (if there are lots of them,
1881 the message gets confusing).
1883 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1884 names, as there are now some PTR records that contain slashes. This check
1885 is only to protect against broken name servers that fall over on strange
1886 characters, so the fact that it applies to all lookups doesn't matter.
1888 PH/36 Now that the new test suite is complete, we can remove some of the
1889 special code in Exim that was needed for the old test suite. For example,
1890 sorting DNS records because real resolvers return them in an arbitrary
1891 order. The new test suite's fake resolver always returns records in the
1894 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1895 bounces) except when queue_only is set, to avoid logging races between
1896 the different processes.
1898 PH/38 Panic-die if .include specifies a non-absolute path.
1900 PH/39 A tweak to the "H" retry rule from its user.
1902 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1903 a label. They prevented compilation on older perls.
1905 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1906 a warning to be raised on newish perls.
1908 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1909 on queue. Changes to match documented behaviour of showing count of
1910 messages matching specified criteria.
1912 PH/40 Changed the default ident timeout from 30s to 5s.
1914 PH/41 Added support for the use of login_cap features, on those BSD systems
1915 that have them, for controlling the resources used by pipe deliveries.
1917 PH/42 The content-scanning code uses fopen() to create files in which to put
1918 message data. Previously it was not paying any attention to the mode of
1919 the files. Exim runs with umask(0) because the rest of the code creates
1920 files with open(), and sets the required mode explicitly. Thus, these
1921 files were ending up world-writeable. This was not a big issue, because,
1922 being within the spool directory, they were not world-accessible. I have
1923 created a function called modefopen, which takes an additional mode
1924 argument. It sets umask(777), creates the file, chmods it to the required
1925 mode, then resets the umask. All the relevant calls to fopen() in the
1926 content scanning code have been changed to use this function.
1928 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1929 to 24 hours. This avoids potential overflow problems when processing G
1930 and H retry rules. I suspect nobody ever tinkers with this value.
1932 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1934 PH/45 When the plaintext authenticator is running as a client, the server's
1935 challenges are checked to ensure they are valid base64 strings. By
1936 default, the authentication attempt is cancelled if an invalid string is
1937 received. Setting client_ignore_invalid_base64 true ignores these errors.
1938 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1939 they are received. Thus, the responses can be made to depend on the
1940 challenges. If an invalid string is ignored, an empty string is placed in
1943 PH/46 Messages that are created by the autoreply transport now contains a
1944 References: header, in accordance with RFCs 2822 and 3834.
1946 PH/47 Added authenticated_sender_force to the smtp transport.
1948 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1950 PH/49 Installed latest patch from the Sieve maintainer.
1952 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1953 was also set, the appendfile transport was unnecessarily scanning a
1954 directory of message files (e.g. for maildir delivery) to find the count
1955 of files (along with the size), even though it did not need this
1956 information. It now does the scan only if it needs to find either the
1957 size of the count of files.
1959 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1961 PH/52 Two bugs concerned with error handling when the smtp transport is
1964 (i) Exim was not creating retry information for temporary errors given
1965 for individual recipients after the DATA command when the smtp transport
1966 was used in LMTP mode. This meant that they could be retried too
1967 frequently, and not timed out correctly.
1969 (ii) Exim was setting the flag that allows error details to be returned
1970 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1971 recipients that were returned after the DATA command.
1973 PH/53 This is related to PH/52, but is more general: for any failing address,
1974 when detailed error information was permitted to be returned to the
1975 sender, but the error was temporary, then after the final timeout, only
1976 "retry timeout exceeded" was returned. Now it returns the full error as
1977 well as "retry timeout exceeded".
1979 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1980 do this, and (what is worse) MTAs that accept it.
1982 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1983 will now be deprecated.
1985 PH/56 New os.c-cygwin from the Cygwin maintainer.
1987 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1988 formats (previously only available in exim formats via -bpr, -bpru,
1989 and -bpra. Now also available in native and exiqgrep formats)
1991 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1992 with very large, slow to parse queues
1994 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1996 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1998 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1999 responses to authentication challenges, though it was showing the
2000 challenges; (ii) I've removed the CR characters from the debug output for
2003 PH/58 Allow for the insertion of a newline as well as a space when a string
2004 is turned into more than one encoded-word during RFC 2047 encoding. The
2005 Sieve code now uses this.
2007 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2008 data_4xx, lost_connection, tls_required.
2010 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2011 message was being sent as an SMTP response.
2013 PH/61 Add -l and -k options to exicyclog.
2015 PH/62 When verifying, if an address was redirected to one new address, so that
2016 verification continued, and the new address failed or deferred after
2017 having set something in $address_data, the value of $address_data was not
2018 passed back to the ACL. This was different to the case when no
2019 redirection occurred. The value is now passed back in both cases.
2021 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2022 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2023 use login_cap.h, so on its own it isn't the distinguishing feature. The
2024 new name refers directly to the setclassresources() function.
2026 PH/65 Added configuration files for NetBSD3.
2028 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2030 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2031 is preferred over IPv4.
2033 PH/68 The bounce_return_message and bounce_return_body options were not being
2034 honoured for bounces generated during the reception of non-SMTP messages.
2035 In particular, this applied to messages rejected by the ACL. This bug has
2036 been fixed. However, if bounce_return_message is true and bounce_return_
2037 body is false, the headers that are returned for a non-SMTP message
2038 include only those that have been read before the error was detected.
2039 (In the case of an ACL rejection, they have all been read.)
2041 PH/69 The HTML version of the specification is now built in a directory called
2042 spec_html instead of spec.html, because the latter looks like a path with
2043 a MIME-type, and this confuses some software.
2045 PH/70 Catch two compiler warnings in sieve.c.
2047 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2048 function verify_get_ident() calls ip_connect() to connect a socket, but
2049 if the "connect()" function timed out, ip_connect() used to close the
2050 socket. However, verify_get_ident() also closes the socket later, and in
2051 between Exim writes to the log, which may get opened at this point. When
2052 the socket was closed in ip_connect(), the log could get the same file
2053 descriptor number as the socket. This naturally causes chaos. The fix is
2054 not to close the socket in ip_connect(); the socket should be closed by
2055 the function that creates it. There was only one place in the code where
2056 this was missing, in the iplookup router, which I don't think anybody now
2057 uses, but I've fixed it anyway.
2059 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2060 well as to direct DNS lookups. Otherwise the handling of names in host
2061 lists is inconsistent and therefore confusing.
2067 PH/01 Two changes to the default runtime configuration:
2069 (1) Move the checks for relay_from_hosts and authenticated clients from
2070 after to before the (commented out) DNS black list checks.
2072 (2) Add control=submission to the relay_from_hosts and authenticated
2073 clients checks, on the grounds that messages accepted by these
2074 statements are most likely to be submissions.
2076 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2078 (1) Generate an error if the third argument for the ${prvs expansion is
2081 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2084 (3) Reset the variables that are obtained from the first argument of
2085 ${prvscheck and used in the second argument before leaving the code,
2086 because their memory is reclaimed, so using them afterwards may do
2089 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2090 one (it's much easier than Tom thought :-).
2092 (5) Because of (4), we can now allow for the use of $prvscheck_result
2093 inside the third argument.
2095 PH/03 For some reason, the default setting of PATH when running a command from
2096 a pipe transport was just "/usr/bin". I have changed it to
2099 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2100 anything to be listed in the output from -bV.
2102 PH/05 When a filter generated an autoreply, the entire To: header line was
2103 quoted in the delivery log line, like this:
2105 => >A.N.Other <ano@some.domain> <original@ddress> ...
2107 This has been changed so that it extracts the operative address. There
2108 may be more than one such address. If so, they are comma-separated, like
2111 => >ano@some.domain,ona@other.domain <original@ddress> ...
2113 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2114 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2115 not being looked up in the rDNS to get a host name, Exim was showing the
2116 IP address twice in Received: lines, even though the IP addresses were
2117 identical. For example:
2119 Received: from [1.2.3.4] (helo=[1.2.3.4])
2121 However, if the real host name was known, it was omitting the HELO data
2122 if it matched the actual IP address. This has been tidied up so that it
2123 doesn't show the same IP address twice.
2125 PH/07 When both +timestamp and +memory debugging was on, the value given by
2126 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2127 called by the debug printing, thereby overwriting the timestamp buffer.
2128 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2130 PH/08 When the original message was included in an autoreply transport, it
2131 always said "this is a copy of the message, including all the headers",
2132 even if body_only or headers_only was set. It now gives an appropriate
2135 PH/09 Applied a patch from the Sieve maintainer which:
2137 o fixes some comments
2138 o adds the (disabled) notify extension core
2139 o adds some debug output for the result of if/elsif tests
2140 o points to the current vacation draft in the documentation
2141 and documents the missing references header update
2145 o fixes a bug in processing the envelope test (when testing
2146 multiple envelope elements, the last element determinted the
2149 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2150 Electronic Mail") by including:
2152 Auto-submitted: auto-generated
2154 in the messages that it generates (bounce messages and others, such as
2155 warnings). In the case of bounce messages for non-SMTP mesages, there was
2156 also a typo: it was using "Auto_submitted" (underscore instead of
2157 hyphen). Since every message generated by Exim is necessarily in response
2158 to another message, thes have all been changed to:
2160 Auto-Submitted: auto-replied
2162 in accordance with these statements in the RFC:
2164 The auto-replied keyword:
2166 - SHOULD be used on messages sent in direct response to another
2167 message by an automatic process,
2169 - MUST NOT be used on manually-generated messages,
2171 - MAY be used on Delivery Status Notifications (DSNs) and Message
2172 Disposition Notifications (MDNs),
2174 - MUST NOT be used on messages generated by automatic or periodic
2175 processes, except for messages which are automatic responses to
2178 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2179 to the default Received: header definition.
2181 PH/12 Added log selector acl_warn_skipped (default on).
2183 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2184 variables because (a) they are in the wrong storage pool and (b) even if
2185 they were copied, it wouldn't work properly because of the caching.
2187 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2188 checking when decoding. Apparently there are clients that generate
2189 overlong encoded strings. Why am I not surprised?
2191 PH/15 If the first argument of "${if match_address" was not empty, but did not
2192 contain an "@" character, Exim crashed. Now it writes a panic log message
2193 and treats the condition as false.
2195 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2197 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2198 extension "envelope-auth". The code is finished and in agreement with
2199 other implementations, but there is no documentation so far and in fact,
2200 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2201 not changing the active code.
2203 Print executed "if" and "elsif" statements when debugging is used. This
2204 helps a great deal to understand what a filter does.
2206 Document more things not specified clearly in RFC3028. I had all this
2207 sorted out, when out of a sudden new issues came to my mind. Oops."
2209 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2212 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2213 canonical form (as documented). However, after a host name lookup from
2214 the IP address, check_host() was doing a simple string comparison with
2215 addresses acquired from the DNS when checking that the found name did
2216 have the original IP as one of its addresses. Since any found IPv6
2217 addresses are likely to be in abbreviated form, the comparison could
2218 fail. Luckily, there already exists a function for doing the comparison
2219 by converting both addresses to binary, so now that is used instead of
2220 the text comparison.
2222 PH/20 There was another similar case to PH/19, when a complete host name was
2223 given in a host list; looking up its IP address could give an abbreviated
2224 form, whereas the current host's name might or might not be abbreviated.
2225 The same fix has been applied.
2231 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2232 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2235 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2236 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2238 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2240 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2241 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2242 Also, for "Auto-Submitted", treat anything other than "no" as
2243 non-personal, in accordance with RFC 3834. (Previously it treated
2244 anything starting "auto-" as non-personal.)
2246 TF/01 The control=submission/name=... option had a problem with syntax
2247 errors if the name included a slash character. The /name= option
2248 now slurps the rest of the string, so it can include any characters
2249 but it must come last in the list of options (after /sender_retain
2252 PH/05 Some modifications to the interface to the fake nameserver for the new
2260 TK/01 Added the "success_on_redirect" address verification option. See
2261 NewStuff for rationale and an example.
2263 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2265 PH/02 Patch to exigrep to allow it to work on syslog lines.
2267 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2268 fread() to skip over the body file's header line, because in Cygwin the
2269 header line is locked and is inaccessible.
2271 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2272 co-exist for some time) to make it clear that it is the Exim ID that is
2273 referenced, not the Message-ID: header line.
2275 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2276 string_format() function, because snprintf() does not exist on all
2279 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2280 ${stat: expansion item.
2282 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2283 protocol synchronization error", to keep the pedants happy.
2285 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2286 well as for IRIX systems, when gcc is being used. See the host.c source
2289 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2291 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2294 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2295 transport and to the smtp transport in LMTP mode.
2297 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2299 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2300 run a filter in a subprocess. This could lead to confusion in subsequent
2301 lookups in the parent process. There should also be a search_tidyup() at
2302 the end of the subprocess.
2304 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2305 only if the host matched helo_try_verify_hosts, which caused the
2306 verification to occur when the EHLO/HELO command was issued. The ACL just
2307 tested the remembered result. Now, if a previous verification attempt has
2308 not happened, "verify = helo" does it there and then.
2310 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2312 TK/03 Fix log output including CR from clamd.
2314 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2315 exist provoked a memory error which could cause a segfault.
2317 PH/15 Installed PCRE 6.2
2319 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2321 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2322 of the problem. Specifically, suggested +O2 rather than +O1 for the
2325 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2327 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2328 record, the port was not correctly logged when the outgoing_port log
2329 selector was set (it logged the transort's default port).
2331 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2332 fallback_hosts, and "hosts" in the smtp transport.
2334 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2335 host errors such as "Connection refused".
2337 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2338 authentication with radiusclient 0.4.9:
2340 - Error returned from rc_read_config was caught wrongly
2341 - Username/password not passed on to radius server due to wrong length.
2343 The presumption is that some radiusclient API changes for 4.51/PH/17
2344 were not taken care of correctly. The code is still untested by me (my
2345 Linux distribution still has 0.3.2 of radiusclient), but it was
2346 contributed by a Radius user.
2348 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2349 expanding the "port" option of the smtp transport.
2351 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2352 while reading a MIME header. Thanks to Tom Hughes for a patch.
2354 PH/24 Include config.h inside local_scan.h so that configuration settings are
2357 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2358 that in an ACL for RCPT (for example), you can examine exactly what was
2361 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2362 commands, but it was not correctly comparing the address with the actual
2363 client host address. Thus, it would show the EHLO address in Received:
2364 header lines when this was not necessary.
2366 PH/27 Added the % operator to ${eval:}.
2368 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2369 it should be ignoring failures (because with -C, for example, it has lost
2370 privilege). It wasn't ignoring creation failures other than "already
2373 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2376 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2377 because an input line was too long, either on its own, or by virtue of
2378 too many continuations, the temporary file was not being removed, and the
2379 return code was incorrect.
2381 PH/31 Missing "BOOL" in function definition in filtertest.c.
2383 PH/32 Applied Sieve patches from the maintainer.
2385 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2387 PH/33 Added "verify = not_blind".
2389 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2390 Local/Makefile (with some defaults set). These are used in built scripts
2391 such as exicyclog, but they have never been used in the exim_install
2392 script (though there are many overriding facilities there). I have
2393 arranged that the exim_install script now takes note of these two
2396 PH/35 Installed configuration files for Dragonfly.
2398 PH/36 When a locally submitted message by a trusted user did not contain a
2399 From: header, and the sender address was obtained from -f or from an SMTP
2400 MAIL command, and the trusted user did not use -F to supply a sender
2401 name, $originator_name was incorrectly used when constructing a From:
2402 header. Furthermore, $originator_name was used for submission mode
2403 messages from external hosts without From: headers in a similar way,
2404 which is clearly wrong.
2406 PH/37 Added control=suppress_local_fixups.
2408 PH/38 When log_selector = +received_sender was set, and the addition of the
2409 sender made the log line's construction buffer exactly full, or one byte
2410 less than full, an overflow happened when the terminating "\n" was
2413 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2414 when the result of a list match is failure because a DNS lookup failed.
2416 PH/40 RM_COMMAND is now used in the building process.
2418 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2419 the "build-* directories that it finds.
2421 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2422 address in a domain literal was a prefix of an interface address.
2424 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2425 when verifying a sender address, unless rewrite_headers is false.
2427 PH/44 Wrote a long comment about why errors_to addresses are verified as
2428 recipients, not senders.
2430 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2431 the ratelimit ACL was added.
2433 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2435 PH/46 Added extra information about PostgreSQL errors to the error string.
2437 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2438 avoiding the need to install special zones in a real server. This is
2439 backwards compatible; if it can't find the fake resolver, it drops back.
2440 Thus, both old and new test suites can be run.
2442 TF/02 Added util/ratelimit.pl
2444 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2445 clock is set back in time.
2447 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2448 Candler <B.Candler@pobox.com>.
2450 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2451 OK for addresses that are the result of redirections.
2453 PH/48 A number of further additions for the benefit of the new test suite,
2454 including a fake gethostbyname() that interfaces to the fake DNS resolver
2457 TF/06 The fix for widen_domains has also been applied to qualify_single and
2458 search_parents which are the other dnslookup options that can cause
2461 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2464 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2466 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2467 service" if it encountered an MX record with an empty target hostname.
2468 The message is now "an MX or SRV record indicated no SMTP service".
2470 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2471 if the DNS of the sending site is misconfigured. This is quite a
2472 common situation. This change restores the behaviour of treating a
2473 helo verification defer as a failure.
2475 PH/51 If self=fail was set on a router, the bounce message did not include the
2476 actual error message.
2482 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2484 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2485 command itself ended in error, the underlying message about the transport
2486 filter timeout was being overwritten with the pipe command error. Now the
2487 underlying error message should be appended to the second error message.
2489 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2491 PH/02 Reduce the amount of output that "make" produces by default. Full output
2492 can still be requested.
2494 PH/03 The warning log line about a condition test deferring for a "warn" verb
2495 was being output only once per connection, rather than after each
2496 occurrence (because it was using the same function as for successful
2497 "warn" verbs). This seems wrong, so I have changed it.
2499 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2500 it should not have, which might have caused a crash in the right
2501 circumstances, but probably never did.
2503 PH/04 Installed a modified version of Tony Finch's patch to make submission
2504 mode fix the return path as well as the Sender: header line, and to
2505 add a /name= option so that you can make the user's friendly name appear
2508 TF/03 Added the control = fakedefer ACL modifier.
2510 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2511 Mark Lowes for thorough testing.
2513 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2515 TK/03 Merged latest SRS patch from Miles Wilton.
2517 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2518 with the definition in sysexits.h (which is #included earlier).
2519 Fortunately, Exim does not actually use EX_OK. The code used to try to
2520 preserve the sysexits.h value, by assumimg that macro definitions were
2521 scanned for macro replacements. I have been disabused of this notion,
2522 so now the code just undefines EX_OK before #including unistd.h.
2524 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2525 in the smtp transport. When a block could not be written in a single
2526 write() function, the timeout was being re-applied to each part-write.
2527 This seems wrong - if the receiver was accepting one byte at a time it
2528 would take for ever. The timeout is now adjusted when this happens. It
2529 doesn't have to be particularly precise.
2531 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2532 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2534 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2535 without a domain if the check to <postmaster@domain> fails.
2537 SC/01 Eximstats: added -xls and the ability to specify output files
2538 (patch written by Frank Heydlauf).
2540 SC/02 Eximstats: use FileHandles for outputing results.
2542 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2544 SC/04 Eximstats: fixed display of large numbers with -nvr option
2546 SC/05 Eximstats: fixed merging of reports with empty tables.
2548 SC/06 Eximstats: added the -include_original_destination flag
2550 SC/07 Eximstats: removed tabs and trailing whitespace.
2552 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2554 TK/06 MBOX spool code: Add real "From " MBOX separator line
2555 so the .eml file is really in mbox format (even though
2556 most programs do not really care). Patch from Alex Miller.
2558 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2559 The latter is generated from $received_to and is only set if the
2560 message has one envelope recipient. SA can use these headers,
2561 obviously out-of-the-box. Patch from Alex Miller.
2563 PH/08 The ${def test on a variable was returning false if the variable's
2564 value was "0", contrary to what the specification has always said!
2565 The result should be true unless the variable is empty.
2567 PH/09 The syntax error of a character other than { following "${if
2568 def:variable_name" (after optional whitespace) was not being diagnosed.
2569 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2570 accidental colon was present, for example, could give incorrect results.
2572 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2573 result is used (not including appendfile, where other changes are about
2576 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2577 This involved changing a lot of size variables from int to off_t. It
2578 should work with maildirs and everything.
2580 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2581 spamd dying while we are connected to it.
2583 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2586 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2587 function for BDB 4.3.
2589 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2591 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2592 the sources for PCRE have been re-arranged and more files are now
2595 PH/15 The code I had for printing potentially long long variables in PH/11
2596 above was not the best (it lost precision). The length of off_t variables
2597 is now inspected at build time, and an appropriate printing format (%ld
2598 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2599 to be "long long int" or "long int". This is needed for the internal
2600 formatting function string_vformat().
2602 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2603 the configuration file to be ":syslog", then the script "guesses" where
2604 the logs files are, rather than using the compiled in default. In our
2605 case the guess is not the same as the compiled default, so the script
2606 suddenly stopped working when I started to use syslog. The patch checks
2607 to see if log_file_path is "". If so, it attempts to read it from exim
2608 with no configuration file to get the compiled in version, before it
2609 falls back to the previous guessing code."
2611 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2612 implementing BATV in an Exim configuration. See NewStuff for the gory
2615 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2616 Makefile that are specific to HP-UX.
2618 PH/18 If the "use_postmaster" option was set for a recipient callout together
2619 with the "random" option, the postmaster address was used as the MAIL
2620 FROM address for the random test, but not for the subsequent recipient
2621 test. It is now used for both.
2623 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2624 patch removes a few documentation additions to RFC 3028, because the
2625 latest draft now contains them. It adds the new en;ascii-case comparator
2626 and a new error check for 8bit text in MIME parts. Comparator and
2627 require names are now matched exactly. I enabled the subaddress
2628 extension, but it is not well tested yet (read: it works for me)."
2630 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2631 rework some of the code of TK/09 above to avoid the hardwired use of
2632 "%lld" and "long long". Replaced the call to snprintf() with a call to
2635 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2636 records point to non-existent hosts", "retry timeout exceeded", and
2637 "retry time not reached for any host after a long failure period".
2639 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2640 experimental DomainKeys support:
2642 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2643 (2) On an error such as an illegally used "control", the wrong name for
2644 the control was given.
2646 These problems did NOT occur unless DomainKeys support was compiled.
2648 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2650 PH/24 Added ${if match_ip condition.
2652 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2653 databases so that it will be absolutely obvious if a crash occurs in the
2654 DB library. This is a regular occurrence (often caused by mis-matched
2657 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2658 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2659 picked up on a user's system that detects such things. There doesn't seem
2660 to be a gcc warning option for this - only an attribute that has to be
2661 put on the function's prototype. It seems that in Fedora Core 4 they have
2662 set this on a number of new functions. No doubt there will be more in due
2665 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2666 specify a transport. However, if an address that was verified by such a
2667 router was the subject of a callout, Exim crashed because it tried to
2668 read the rcpt_include_affixes from the non-existent transport. Now it
2669 just assumes that the setting of that option is false. This bug was
2670 introduced by 4.51/PH/31.
2672 PH/28 Changed -d+all to exclude +memory, because that information is very
2673 rarely of interest, but it makes the output a lot bigger. People tend to
2674 do -d+all out of habit.
2676 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2677 code in os-type was giving problems when libc.so lives in lib64, like on
2680 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2681 aren't the modern standard, and it seems that some systems' include files
2682 don't always have them. Exim was already checking for some of the newer
2683 ones like T_AAAA, and defining it itself. I've added checks for all the
2684 record types that Exim uses.
2686 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2687 not automatically generating a new one, as it is supposed to. This
2688 prevented TLS from working. If the file did exist, but contained invalid
2689 data, a new version was generated, as expected. It was only the case of a
2690 non-existent file that was broken.
2692 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2693 with a change in libdomainkeys > 0.64.
2695 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2696 from DNS. If the selector record carries the flag, it now has
2697 precedence over the domain-wide flag.
2699 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2701 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2702 the use of a port name that isn't defined in /etc/services) occurred, the
2703 message was deferred as in a normal delivery, and thus remained on the
2704 spool, instead of being failed because of the mua_wrapper setting. This
2705 is now fixed, and I tidied up some of the mua_wrapper messages at the
2708 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2709 the messages in a hash of arrays rather than using individual hashes.
2710 This is a bit cleaner and results in dramatic memory savings, albeit
2711 at a slight CPU cost.
2713 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2714 as requested by Marc Sherman.
2716 SC/10 Eximstats: added histograms for user specified patterns as requested
2719 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2721 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2722 fopen() in the content-scanning modules that did not already have it.
2728 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2729 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2731 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2733 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2735 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2736 to test Sieve filters that use "vacation".
2738 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2739 that changes the way the GnuTLS parameters are stored in the cache file.
2740 The new format can be generated externally. For backward compatibility,
2741 if the data in the cache doesn't make sense, Exim assumes it has read an
2742 old-format file, and it generates new data and writes a new file. This
2743 means that you can't go back to an older release without removing the
2746 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2747 work if there are any delivery delays because "one_time" forces the
2748 parent to be marked "delivered", so its unseen clone is never tried
2749 again. For this reason, Exim now forbids the simultaneous setting of
2752 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2753 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2754 entry if you want to know the details. The fix, however, neglected to
2755 consider the case when local delivery batching is involved. The test for
2756 "previously delivered" was not happening when checking to see if an
2757 address could be batched with a previous (undelivered) one; under
2758 certain circumstances this could lead to multiple deliveries to the same
2761 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2762 in its include files, and this causes problems building Exim.
2764 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2765 header_syntax) but Exim was just ignoring anything given after a slash.
2766 In particular, this caused confusion with an attempt to use "verify =
2767 reverse_host_lookup/defer_ok". An error is now given when options are
2768 supplied for verify items that do not have them. (Maybe reverse_host_
2769 lookup should have a defer_ok option, but that's a different point.)
2771 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2772 defined by RFC 821) to 2048, because there were problems with some AUTH
2773 commands, and RFC 1869 says the size should be increased for extended
2774 SMTP commands that take arguments.
2776 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2779 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2780 "unknown" error; now it says that the functionality isn't in the binary.
2782 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2783 an address' error message when a string expansion fails (syntax or
2784 whatever). Otherwise the password may appear in the log. Following change
2785 PH/42 below, there is no longer a chance of it appearing in a bounce
2788 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2790 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
2791 of its addresses was ever tried. (Bugzilla bug #2.)
2793 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2794 the result incorrectly in the debug output. (It correctly added a newline
2795 to what was transported.)
2797 TF/01 Added $received_time.
2799 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2800 commented out examples of how to interface to a virus scanner and to
2801 SpamAssassin. Also added commented examples of av_scanner and
2802 spamd_address settings.
2804 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2805 and controls are allowed in which ACLs. There were a couple of minor
2806 errors. Some of the entries in the conditions table (which is a table of
2807 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2808 negation of where the condition IS allowed.
2810 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2812 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2813 header file does not have a version number, so I've had to invent a new
2814 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2815 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2816 radiusclient), but it was contributed by a Radius user.
2818 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2819 files or directories, for OpenSSL.
2821 PH/19 When an Exim process that is running as root has to create an Exim log
2822 file, it does so in a subprocess that runs as exim:exim so as to get the
2823 ownership right at creation (otherwise, other Exim processes might see
2824 the file with the wrong ownership). There was no test for failure of this
2825 fork() call, which would lead to the process getting stuck as it waited
2826 for a non-existent subprocess. Forks do occasionally fail when resources
2827 run out. I reviewed all the other calls to fork(); they all seem to check
2830 PH/20 When checking for unexpected SMTP input at connect time (before writing
2831 the banner), Exim was not dealing correctly with a non-positive return
2832 from the read() function. If the client had disconnected by this time,
2833 the result was a log entry for a synchronization error with an empty
2834 string after "input=" when read() returned zero. If read() returned -1
2835 (an event I could not check), uninitialized data bytes were printed.
2836 There were reports of junk text (parts of files, etc) appearing after
2839 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2841 PH/22 Added support for macro redefinition, and (re)definition in between
2842 driver and ACL definitions.
2844 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2845 forgetting to use the resulting value; it was using the unexpanded value.
2847 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2848 hadn't been configured. The fix is from Juergen Kreileder, who
2849 understands it better than I do:
2851 "Here's what I see happening with three configured cyrus_sasl
2852 authenticators configured (plain, login, cram-md5):
2854 On startup auth_cyrus_sasl_init() gets called for each of these.
2855 This means three calls to sasl_listmech() without a specified mech_list.
2856 => SASL tests which mechs of all available mechs actually work
2857 => three warnings about OTP not working
2858 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2860 With the patch, sasl_listmech() also gets called three times. But now
2861 SASL's mech_list option is set to the server_mech specified in the the
2862 authenticator. Or in other words, the answer from sasl_listmech()
2863 gets limited to just the mech you're testing for (which is different
2865 => the return list contains just 'plain' or 'login', 'cram-md5' or
2866 nothing depending on the value of ob->server_mech.
2868 I've just tested the patch: Authentication still works fine,
2869 unavailable mechs specified in the exim configuration are still
2870 caught, and the auth.log warnings about OTP are gone."
2872 PH/25 When debugging is enabled, the contents of the command line are added
2873 to the debugging output, even when log_selector=+arguments is not
2876 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2877 answer is "GNU", and only if the return is "GNU/something" is the answer
2880 PH/27 $acl_verify_message is now set immediately after the failure of a
2881 verification in an ACL, and so is available in subsequent modifiers. In
2882 particular, the message can be preserved by coding like this:
2884 warn !verify = sender
2885 set acl_m0 = $acl_verify_message
2887 Previously, $acl_verify_message was set only while expanding "message"
2888 and "log_message" when a very denied access.
2890 PH/28 Modified OS/os.c-Linux with
2892 -#ifndef OS_LOAD_AVERAGE
2893 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2895 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2896 nomenclature these days.)
2898 PH/29 Installed patch from the Sieve maintainer that adds the options
2899 sieve_useraddress and sieve_subaddress to the redirect router.
2901 PH/30 In these circumstances:
2902 . Two addresses routed to the same list of hosts;
2903 . First host does not offer TLS;
2904 . First host accepts first address;
2905 . First host gives temporary error to second address;
2906 . Second host offers TLS and a TLS session is established;
2907 . Second host accepts second address.
2908 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2909 and peerdn, if requested) that were in fact used only for the second
2912 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2913 attention to any local part prefix or suffix that was matched by the
2914 router that accepted the address. It now behaves in the same way as it
2915 does for delivery: the affixes are removed from the local part unless
2916 rcpt_include_affixes is set on the transport.
2918 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2919 timeout during the DATA phase of an incoming message.
2921 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2922 applied a patch sanctioned by the Sieve maintainer.
2924 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2925 the uid or gid is negative. A case of a negative gid caused this to be
2926 noticed. The fix allows for either to be negative.
2928 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2929 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2932 PH/36 The change PH/12 above was broken. Fixed it.
2934 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2935 the grounds that routing the same address twice would always produce the
2936 same answer. This might have been true once, but it is certainly no
2937 longer true now. Routing a child address may depend on the previous
2938 routing that produced that child. Some complicated redirection strategies
2939 went wrong when messages had multiple recipients, and made Exim's
2940 behaviour dependent on the order in which the addresses were given.
2942 I have moved the duplicate checking until after the routing is complete.
2943 Exim scans the addresses that are assigned to local and remote
2944 transports, and removes any duplicates. This means that more work will be
2945 done, as duplicates will always all be routed, but duplicates are
2946 presumably rare, so I don't expect this is of any significance.
2948 For deliveries to pipes, files, and autoreplies, the duplicate checking
2949 still happens during the routing process, since they are not going to be
2952 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2953 It corrects a timeout issue with spamd. This is Ian's comment: "The
2954 background is that sometimes spamd either never reads data from a
2955 connection it has accepted, or it never writes response data. The exiscan
2956 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2957 blindly assumes that writes won't block so it may never time out."
2959 PH/39 Allow G after quota size as well as K and M.
2961 PH/40 The value set for $authenticated_id in an authenticator may not contain
2962 binary zeroes or newlines because the value is written to log lines and
2963 to spool files. There was no check on this. Now the value is run through
2964 the string_printing() function so that such characters are converted to
2965 printable escape sequences.
2967 PH/41 $message_linecount is a new variable that contains the total number of
2968 lines in the message. Compare $body_linecount, which is the count for the
2971 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2972 bounce and delay warning messages, except in certain special cases, which
2975 (a) An SMTP error message from a remote host;
2976 (b) A message specified in a :fail: redirection;
2977 (c) A message specified in a "fail" command in a system filter;
2978 (d) A message specified in a FAIL return from the queryprogram router;
2979 (e) A message specified by the cannot_route_message router option.
2981 In these cases only, Exim does include the error details in bounce and
2982 warning messages. There are also a few cases where bland messages such
2983 as "unrouteable address" or "local delivery error" are given.
2985 PH/43 $value is now also set for the "else" part of a ${run expansion.
2987 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2988 being worked on, but at least Exim now implements the latest version to
2991 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2992 process to complete was treated as a delivery failure, a timeout while
2993 writing the message to the pipe was logged, but erroneously treated as a
2994 successful delivery. Such timeouts include transport filter timeouts. For
2995 consistency with the overall process timeout, these timeouts are now
2996 treated as errors, giving rise to delivery failures by default. However,
2997 there is now a new Boolean option for the pipe transport called
2998 timeout_defer, which, if set TRUE, converts the failures into defers for
2999 both kinds of timeout. A transport filter timeout is now identified in
3002 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3003 systems where "make" and "gmake" are different, calling "gmake" at top
3004 level broke things. I've arranged for the value of $(MAKE) to be passed
3005 from the Makefile to this script so that it can call the same version of
3009 A note about Exim versions 4.44 and 4.50
3010 ----------------------------------------
3012 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3013 changes of various kinds. As a consequence, a big documentation update was
3014 needed. This delayed the release for rather longer than seemed good, especially
3015 in the light of a couple of (minor) security issues. Therefore, the changes
3016 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3017 release. So 4.44 and 4.50 are in effect two different branches that both start
3020 I have left the 4.50 change log unchanged; it contains all the changes since
3021 4.43. The change log for 4.44 is below; many of its items are identical to
3022 those for 4.50. This seems to be the most sensible way to preserve the
3023 historical information.
3029 1. Minor wording change to the doc/README.SIEVE file.
3031 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3032 computation of the current number of files was incorrect.
3034 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3035 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3038 4. Give more explanation in the error message when the command for a transport
3039 filter fails to execute.
3041 5. There are several places where Exim runs a non-Exim command in a
3042 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3043 was being done only for the command run by the queryprogram router. It is
3044 now done for all such subprocesses. The other cases are: ${run, transport
3045 filters, and the commands run by the lmtp and pipe transports.
3047 6. Added CONFIGURE_GROUP build-time option.
3049 7. Some older OS have a limit of 256 on the maximum number of file
3050 descriptors. Exim was using setrlimit() to set 1000 as a large value
3051 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3052 systems. I've change it so that if it can't get 1000, it tries for 256.
3054 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3055 was an oversight, and furthermore, ever since the addition of extra
3056 controls (e.g. 4.43/32), the checks on when to allow different forms of
3057 "control" were broken. There should now be diagnostics for all cases when a
3058 control that does not make sense is encountered.
3060 9. Added the /retain_sender option to "control=submission".
3062 10. $recipients is now available in the predata ACL (oversight).
3064 11. Tidy the search cache before the fork to do a delivery from a message
3065 received from the command line. Otherwise the child will trigger a lookup
3066 failure and thereby defer the delivery if it tries to use (for example) a
3067 cached ldap connection that the parent has called unbind on.
3069 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3070 of $address_data from the recipient verification was clobbered by the
3071 sender verification.
3073 13. The value of address_data from a sender verification is now available in
3074 $sender_address_data in subsequent conditions in the ACL statement.
3076 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3078 15. Added a new option "connect=<time>" to callout options, to set a different
3081 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3082 was its contents. (It was OK if the option was not defined at all.)
3084 17. A "Completed" log line is now written for messages that are removed from
3085 the spool by the -Mrm option.
3087 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3088 information about exactly what failed.
3090 19. Added -dd to debug only the daemon process.
3092 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3093 handles timeouts, both on the server side and network timeouts. Renamed the
3094 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3096 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3097 It is now set to "smtps".
3099 22. $host_address is now set to the target address during the checking of
3100 ignore_target_hosts.
3102 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3103 being passed; this would have caused $sender_host_name to have been used if
3104 matching the list had actually called for a host name (not very likely,
3105 since this list is usually IP addresses). A host name is now passed as
3108 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3109 code that specifies a non-restarting handler (typically sigaction() in
3110 modern systems) in an attempt to fix a rare and obscure crash bug.
3112 25. Narrowed the window for a race in the daemon that could cause it to ignore
3113 SIGCHLD signals. This is not a major problem, because they are used only to
3114 wake it up if nothing else does.
3116 26. A malformed maildirsize file could cause Exim to calculate negative values
3117 for the mailbox size or file count. Odd effects could occur as a result.
3118 The maildirsize information is now recalculated if the size or filecount
3121 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3122 support for a long time. Removed HAVE_SYS_VFS_H.
3124 28. Installed the latest version of exipick from John Jetmore.
3126 29. In an address list, if the pattern was not a regular expression, an empty
3127 subject address (from a bounce message) matched only if the pattern was an
3128 empty string. Non-empty patterns were not even tested. This was the wrong
3129 because it is perfectly reasonable to use an empty address as part of a
3130 database query. An empty address is now tested by patterns that are
3131 lookups. However, all the other forms of pattern expect the subject to
3132 contain a local part and a domain, and therefore, for them, an empty
3133 address still always fails if the pattern is not itself empty.
3135 30. Exim went into a mad DNS loop when attempting to do a callout where the
3136 host was specified on an smtp transport, and looking it up yielded more
3137 than one IP address.
3139 31. Re-factored the code for checking spool and log partition space into a
3140 function that finds that data and another that does the check. The former
3141 is then used to implement four new variables: $spool_space, $log_space,
3142 $spool_inodes, and $log_inodes.
3144 32. The RFC2047 encoding function was originally intended for short strings
3145 such as real names; it was not keeping to the 75-character limit for
3146 encoded words that the RFC imposes. It now respects the limit, and
3147 generates multiple encoded words if necessary. To be on the safe side, I
3148 have increased the buffer size for the ${rfc2047: expansion operator from
3151 33. It is now permitted to omit both strings after an "if" condition; if the
3152 condition is true, the result is "true". As before, when the second string
3153 is omitted, a false condition yields an empty string. This makes it less
3154 cumbersome to write custom ACL and router conditions.
3156 34. Failure to deliver a bounce message always caused it to be frozen, even if
3157 there was an errors_to setting on the router. The errors_to setting is now
3160 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3161 canonical form (fully expanded) before being placed in
3162 $sender_host_address.
3164 36. The table in the code that translates DNS record types into text (T_A to
3165 "A" for instance) was missing entries for NS and CNAME. It is just possible
3166 that this could have caused confusion if both these types were looked up
3167 for the same domain, because the text type is used as part of Exim's
3168 per-process caching. But the chance of anyone hitting this buglet seems
3171 37. The dnsdb lookup has been extended in a number of ways.
3173 (1) There is a new type, "zns", which walks up the domain tree until it
3174 finds some nameserver records. It should be used with care.
3176 (2) There is a new type, "mxh", which is like "mx" except that it returns
3177 just the host names, not the priorities.
3179 (3) It is now possible to give a list of domains (or IP addresses) to be
3180 looked up. The behaviour when one of the lookups defers can be
3181 controlled by a keyword.
3183 (4) It is now possible to specify the separator character for use when
3184 multiple records are returned.
3186 38. The dnslists ACL condition has been extended: it is now possible to supply
3187 a list of IP addresses and/or domains to be looked up in a particular DNS
3190 39. Added log_selector=+queue_time_overall.
3192 40. When running the queue in the test harness, wait just a tad after forking a
3193 delivery process, to get repeatability of debugging output.
3195 41. Include certificate and key file names in error message when GnuTLS fails
3196 to set them up, because the GnuTLS error message doesn't include the name
3197 of the failing file when there is a problem reading it.
3199 42. Allow both -bf and -bF in the same test run.
3201 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3203 44. The "Exiscan patch" is now merged into the mainline Exim source.
3205 45. Sometimes the final signoff response after QUIT could fail to get
3206 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3207 < 0 before doing a fflush(). This bug looks as though it goes back to the
3208 introduction of TLS in release 3.20, but "sometimes" must have been rare
3209 because the tests only now provoked it.
3211 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3212 (this can affect the format of dates).
3214 47. exim_tidydb, when checking for the continued existence of a message for
3215 which it has found a message-specific retry record, was not finding
3216 messages that were in split spool directories. Consequently, it was
3217 deleting retry records that should have stayed in existence.
3219 48. Steve fixed some bugs in eximstats.
3221 49. The SPA authentication driver was not abandoning authentication and moving
3222 on to the next authenticator when an expansion was forced to fail,
3223 contradicting the general specification for all authenticators. Instead it
3224 was generating a temporary error. It now behaves as specified.
3226 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3227 (the order specifies the preference for clients). The order is now AES256,
3228 AES128, 3DES, ARCFOUR128.
3230 51. Small patch to Sieve code - explicitly set From: when generating an
3233 52. Exim crashed if a remote delivery caused a very long error message to be
3234 recorded - for instance if somebody sent an entire SpamAssassin report back
3235 as a large number of 550 error lines. This bug was coincidentally fixed by
3236 increasing the size of one of Exim's internal buffers (big_buffer) that
3237 happened as part of the Exiscan merge. However, to be on the safe side, I
3238 have made the code more robust (and fixed the comments that describe what
3241 53. Now that there can be additional text after "Completed" in log lines (if
3242 the queue_time_overall log selector is set), a one-byte patch to exigrep
3243 was needed to allow it to recognize "Completed" as not the last thing in
3246 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3247 patch that reportedly fixes this has been added. I am not expert enough to
3248 create a test for it. This is what the patch creator wrote:
3250 "I found a little strange behaviour of ldap code when working with
3251 Windows 2003 AD Domain, where users was placed in more than one
3252 Organization Units. When I tried to give exim partial DN, the exit code
3253 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3254 But simultaneously result of request was absolutely normal ldap result,
3255 so I produce this patch..."
3257 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3258 REFERENCE, so I have modified the code to exclude the patch when that macro
3261 55. Some experimental protocols are using DNS PTR records for new purposes. The
3262 keys for these records are domain names, not reversed IP addresses. The
3263 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3264 leaves it alone. Component reversal etc. now happens only for IP addresses.
3267 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3269 57. Double the size of the debug message buffer (to 2048) so that more of very
3270 long debug lines gets shown.
3272 58. The exicyclog utility now does better if the number of log files to keep
3273 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3275 59. Two changes related to the smtp_active_hostname option:
3277 (1) $smtp_active_hostname is now available as a variable.
3278 (2) The default for smtp_banner uses $smtp_active_hostname instead
3279 of $primary_hostname.
3281 60. The host_aton() function is supposed to be passed a string that is known
3282 to be a valid IP address. However, in the case of IPv6 addresses, it was
3283 not checking this. This is a hostage to fortune. Exim now panics and dies
3284 if the condition is not met. A case was found where this could be provoked
3285 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3286 components; fortuitously, this particular loophole had already been fixed
3287 by change 4.50/55 above.
3289 If there are any other similar loopholes, the new check in host_aton()
3290 itself should stop them being exploited. The report I received stated that
3291 data on the command line could provoke the exploit when Exim was running as
3292 exim, but did not say which command line option was involved. All I could
3293 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3294 running as the user.
3297 61. There was a buffer overflow vulnerability in the SPA authentication code
3298 (which came originally from the Samba project). I have added a test to the
3299 spa_base64_to_bits() function which I hope fixes it.
3302 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3303 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3305 63. The daemon start-up calls getloadavg() while still root for those OS that
3306 need the first call to be done as root, but it missed one case: when
3307 deliver_queue_load_max is set with deliver_drop_privilege. This is
3308 necessary for the benefit of the queue runner, because there is no re-exec
3309 when deliver_drop_privilege is set.
3311 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3312 This has been fixed.
3314 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3315 was in use, was not putting the data itself into the right store pool;
3316 consequently, it could be overwritten for a subsequent message in the same
3317 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3320 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3322 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3323 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3324 the function were treating the return as a boolean value, which happened to
3325 work because 0=false and not-0=true, but is not correct code.
3327 68. The host_aton() function was not handling scoped IPv6 addresses (those
3328 with, for example, "%eth0" on the end) correctly.
3330 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3331 negated items (that is, ~something) in unsigned ints. Some compilers
3332 apparently mutter when there is no cast.
3334 70. If an address verification called from an ACL failed, and did not produce a
3335 user-specific message (i.e. there was only a "system" message), nothing was
3336 put in $acl_verify_message. In this situation, it now puts the system
3339 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3340 session; change 4.31/43 added the unwanted input to the log line - except
3341 that it did not do this in the start of session case. It now does.
3343 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3344 This is wrong and can cause the other end to generate a synchronization
3345 error if it is another Exim or anything else that does the synchronization
3346 check. A QUIT command is no longer sent after a timeout.
3348 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3349 during host lookups.
3351 74. The defer_ok option of callout verification was not working if it was used
3352 when verifying addresses in header lines, that is, for this case:
3354 verify = header_sender/callout=defer_ok
3356 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3357 those file descriptors could be used for SMTP connections. If anything
3358 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3359 could be sent to the SMTP client, causing chaos. The daemon now opens
3360 stdin, stdout, and stderr to /dev/null when it puts itself into the
3363 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3364 log by default. The user can override this with suitable Perl magic.
3366 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3367 the log message when discard triggers, was not working for the DATA ACL or
3368 for the non-SMTP ACL.
3370 78. Error message wording change in sieve.c.
3372 79. If smtp_accept_max_per_host was set, the number of connections could be
3373 restricted to fewer than expected, because the daemon was trying to set up
3374 a new connection before checking whether the processes handling previous
3375 connections had finished. The check for completed processes is now done
3376 earlier. On busy systems, this bug wouldn't be noticed because something
3377 else would have woken the daemon, and it would have reaped the completed
3380 80. If a message was submitted locally by a user whose login name contained one
3381 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3382 It caused a spool format error. I have fixed the spool reading code. A
3383 related problem was that the "from" clause in the Received: line became
3384 illegal because of the space(s). It is now covered by ${quote_local_part.
3386 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3389 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3391 83. Patch from Sieve maintainer to fix unterminated string problem in
3392 "vacation" handling.
3394 84. Some minor changes to the Linux configuration files to help with other
3395 OS variants using glibc.
3397 85. One more patch for Sieve to update vacation handling to latest spec.
3400 ----------------------------------------------------
3401 See the note above about the 4.44 and 4.50 releases.
3402 ----------------------------------------------------
3408 1. Change 4.43/35 introduced a bug that caused file counts to be
3409 incorrectly computed when quota_filecount was set in an appendfile
3412 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3413 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3416 3. Give more explanation in the error message when the command for a transport
3417 filter fails to execute.
3419 4. There are several places where Exim runs a non-Exim command in a
3420 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3421 was being done only for the command run by the queryprogram router. It is
3422 now done for all such subprocesses. The other cases are: ${run, transport
3423 filters, and the commands run by the lmtp and pipe transports.
3425 5. Some older OS have a limit of 256 on the maximum number of file
3426 descriptors. Exim was using setrlimit() to set 1000 as a large value
3427 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3428 systems. I've change it so that if it can't get 1000, it tries for 256.
3430 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3431 was an oversight, and furthermore, ever since the addition of extra
3432 controls (e.g. 4.43/32), the checks on when to allow different forms of
3433 "control" were broken. There should now be diagnostics for all cases when a
3434 control that does not make sense is encountered.
3436 7. $recipients is now available in the predata ACL (oversight).
3438 8. Tidy the search cache before the fork to do a delivery from a message
3439 received from the command line. Otherwise the child will trigger a lookup
3440 failure and thereby defer the delivery if it tries to use (for example) a
3441 cached ldap connection that the parent has called unbind on.
3443 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3444 of $address_data from the recipient verification was clobbered by the
3445 sender verification.
3447 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3448 was its contents. (It was OK if the option was not defined at all.)
3450 11. A "Completed" log line is now written for messages that are removed from
3451 the spool by the -Mrm option.
3453 12. $host_address is now set to the target address during the checking of
3454 ignore_target_hosts.
3456 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3457 being passed; this would have caused $sender_host_name to have been used if
3458 matching the list had actually called for a host name (not very likely,
3459 since this list is usually IP addresses). A host name is now passed as
3462 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3463 code that specifies a non-restarting handler (typically sigaction() in
3464 modern systems) in an attempt to fix a rare and obscure crash bug.
3466 15. Narrowed the window for a race in the daemon that could cause it to ignore
3467 SIGCHLD signals. This is not a major problem, because they are used only to
3468 wake it up if nothing else does.
3470 16. A malformed maildirsize file could cause Exim to calculate negative values
3471 for the mailbox size or file count. Odd effects could occur as a result.
3472 The maildirsize information is now recalculated if the size or filecount
3475 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3476 support for a long time. Removed HAVE_SYS_VFS_H.
3478 18. Updated exipick to current release from John Jetmore.
3480 19. Allow an empty sender to be matched against a lookup in an address list.
3481 Previously the only cases considered were a regular expression, or an
3484 20. Exim went into a mad DNS lookup loop when doing a callout where the
3485 host was specified on the transport, if the DNS lookup yielded more than
3488 21. The RFC2047 encoding function was originally intended for short strings
3489 such as real names; it was not keeping to the 75-character limit for
3490 encoded words that the RFC imposes. It now respects the limit, and
3491 generates multiple encoded words if necessary. To be on the safe side, I
3492 have increased the buffer size for the ${rfc2047: expansion operator from
3495 22. Failure to deliver a bounce message always caused it to be frozen, even if
3496 there was an errors_to setting on the router. The errors_to setting is now
3499 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3500 canonical form (fully expanded) before being placed in
3501 $sender_host_address.
3503 24. Updated eximstats to version 1.33
3505 25. Include certificate and key file names in error message when GnuTLS fails
3506 to set them up, because the GnuTLS error message doesn't include the name
3507 of the failing file when there is a problem reading it.
3509 26. Expand error message when OpenSSL has problems setting up cert/key files.
3512 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3513 (this can affect the format of dates).
3515 28. exim_tidydb, when checking for the continued existence of a message for
3516 which it has found a message-specific retry record, was not finding
3517 messages that were in split spool directories. Consequently, it was
3518 deleting retry records that should have stayed in existence.
3520 29. eximstats updated to version 1.35
3521 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3522 1.35 - bugfix such that pie charts by volume are generated correctly
3524 30. The SPA authentication driver was not abandoning authentication and moving
3525 on to the next authenticator when an expansion was forced to fail,
3526 contradicting the general specification for all authenticators. Instead it
3527 was generating a temporary error. It now behaves as specified.
3529 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3530 (the order specifies the preference for clients). The order is now AES256,
3531 AES128, 3DES, ARCFOUR128.
3533 31. Small patch to Sieve code - explicitly set From: when generating an
3536 32. Exim crashed if a remote delivery caused a very long error message to be
3537 recorded - for instance if somebody sent an entire SpamAssassin report back
3538 as a large number of 550 error lines. This bug was coincidentally fixed by
3539 increasing the size of one of Exim's internal buffers (big_buffer) that
3540 happened as part of the Exiscan merge. However, to be on the safe side, I
3541 have made the code more robust (and fixed the comments that describe what
3544 33. Some experimental protocols are using DNS PTR records for new purposes. The
3545 keys for these records are domain names, not reversed IP addresses. The
3546 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3547 leaves it alone. Component reversal etc. now happens only for IP addresses.
3550 34. The host_aton() function is supposed to be passed a string that is known
3551 to be a valid IP address. However, in the case of IPv6 addresses, it was
3552 not checking this. This is a hostage to fortune. Exim now panics and dies
3553 if the condition is not met. A case was found where this could be provoked
3554 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3555 components; fortuitously, this particular loophole had already been fixed
3556 by change 4.50/55 or 4.44/33 above.
3558 If there are any other similar loopholes, the new check in host_aton()
3559 itself should stop them being exploited. The report I received stated that
3560 data on the command line could provoke the exploit when Exim was running as
3561 exim, but did not say which command line option was involved. All I could
3562 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3563 running as the user.
3566 35. There was a buffer overflow vulnerability in the SPA authentication code
3567 (which came originally from the Samba project). I have added a test to the
3568 spa_base64_to_bits() function which I hope fixes it.
3571 36. The daemon start-up calls getloadavg() while still root for those OS that
3572 need the first call to be done as root, but it missed one case: when
3573 deliver_queue_load_max is set with deliver_drop_privilege. This is
3574 necessary for the benefit of the queue runner, because there is no re-exec
3575 when deliver_drop_privilege is set.
3577 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3578 was in use, was not putting the data itself into the right store pool;
3579 consequently, it could be overwritten for a subsequent message in the same
3580 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3583 38. Sometimes the final signoff response after QUIT could fail to get
3584 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3585 < 0 before doing a fflush(). This bug looks as though it goes back to the
3586 introduction of TLS in release 3.20, but "sometimes" must have been rare
3587 because the tests only now provoked it.
3593 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3594 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3595 Now it returns an integer. A number of calls were still expecting a T/F
3596 return. Fortuitously, in all cases, the tests worked in OK situations,
3597 which is the norm. However, things would have gone wrong on any write
3598 failures on the smtp file descriptor. This function is used when sending
3599 messages over SMTP and also when doing verify callouts.
3601 2. When Exim is called to do synchronous delivery of a locally submitted
3602 message (the -odf or -odi options), it no longer closes stderr before doing
3605 3. Implemented the mua_wrapper option.
3607 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3609 5. Implemented the functions header_remove(), header_testname(),
3610 header_add_at_position(), and receive_remove_recipient(), and exported them
3613 6. If an ACL "warn" statement specified the addition of headers, Exim already
3614 inserted X-ACL-Warn: at the start if there was no header name. However, it
3615 was not making this test for the second and subsequent header lines if
3616 there were newlines in the string. This meant that an invalid header could
3617 be inserted if Exim was badly configured.
3619 7. Allow an ACL "warn" statement to add header lines at the start or after all
3620 the Received: headers, as well as at the end.
3622 8. Added the rcpt_4xx retry error code.
3624 9. Added postmaster_mailfrom=xxx to callout verification option.
3626 10. Added mailfrom=xxxx to the callout verification option, for verify=
3629 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3630 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3632 12. Inserted some casts to stop certain compilers warning when using pointer
3633 differences as field lengths or precisions in printf-type calls (mostly
3634 affecting debugging statements).
3636 13. Added optional readline() support for -be (dynamically loaded).
3638 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3639 same clock tick as a message's arrival, so that its received time was the
3640 same as the "first fail" time on the retry record, and that message
3641 remained on the queue past the ultimate address timeout, every queue runner
3642 would try a delivery (because it was past the ultimate address timeout) but
3643 after another failure, the ultimate address timeout, which should have then
3644 bounced the address, did not kick in. This was a "< instead of <=" error;
3645 in most cases the first failure would have been in the next clock tick
3646 after the received time, and all would be well.
3648 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3649 being recognized when the domain list was tested by the match_domain
3650 condition in an expansion string.
3652 16. Added the ${str2b64: operator.
3654 17. Exim was always calling setrlimit() to set a large limit for the number of
3655 processes, without checking whether the existing limit was already
3656 adequate. (It did check for the limit on file descriptors.) Furthermore,
3657 errors from getrlimit() and setrlimit() were being ignored. Now they are
3658 logged to the main and panic logs, but Exim does carry on, to try to do its
3659 job under whatever limits there are.
3661 18. Imported PCRE 5.0.
3663 19. Trivial typo in log message " temporarily refused connection" (the leading
3666 20. If the log selector return_path_on_delivery was set and an address was
3667 redirected to /dev/null, the delivery process crashed because it assumed
3668 that a return path would always be set for a "successful" delivery. In this
3669 case, the whole delivery is bypassed as an optimization, and therefore no
3672 21. Internal re-arrangement: the function for sending a challenge and reading
3673 a response while authentication was assuming a zero-terminated challenge
3674 string. It's now changed to take a pointer and a length, to allow for
3675 binary data in such strings.
3677 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3679 23. Exim was not respecting finduser_retries when seeking the login of the
3680 uid under which it was called; it was always trying 10 times. (The default
3681 setting of finduser_retries is zero.) Also, it was sleeping after the final
3682 failure, which is pointless.
3684 24. Implemented tls_on_connect_ports.
3686 25. Implemented acl_smtp_predata.
3688 26. If the domain in control=submission is set empty, Exim assumes that the
3689 authenticated id is a complete email address when it generates From: or
3690 Sender: header lines.
3692 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3693 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3694 chown and chgrp in /bin and hostname in /usr/bin.
3696 28. Exim was keeping the "process log" file open after each use, just as it
3697 does for the main log. This opens the possibility of it remaining open for
3698 long periods when the USR1 signal hits a daemon. Occasional processlog
3699 errors were reported, that could have been caused by this. Anyway, it seems
3700 much more sensible not to leave this file open at all, so that is what now
3703 29. The long-running daemon process does not normally write to the log once it
3704 has entered its main loop, and it closes the log before doing so. This is
3705 so that log files can straightforwardly be renamed and moved. However,
3706 there are a couple of unusual error situations where the daemon does write
3707 log entries, and I had neglected to close the log afterwards.
3709 30. The text of an SMTP error response that was received during a remote
3710 delivery was being truncated at 512 bytes. This is too short for some of
3711 the long messages that one sometimes sees. I've increased the limit to
3714 31. It is now possible to make retry rules that apply only when a message has a
3715 specific sender, in particular, an empty sender.
3717 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3718 it possible to be selective about when SMTP synchronization is enforced.
3720 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3722 32. Implemented hosts_connection_nolog.
3724 33. Added an ACL for QUIT.
3726 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3729 35. Added mailbox_size and mailbox_filecount to appendfile.
3731 36. Added control = no_multiline_responses to ACLs.
3733 37. There was a bug in the logic of the code that waits for the clock to tick
3734 in the case where the clock went backwards by a substantial amount such
3735 that the microsecond fraction of "now" was more than the microsecond
3736 fraction of "then" (but the whole seconds number was less).
3738 38. Added support for the libradius Radius client library this is found on
3739 FreeBSD (previously only the radiusclient library was supported).
3745 1. When certain lookups returned multiple values in the form name=value, the
3746 quoting of the values was not always being done properly. Specifically:
3747 (a) If the value started with a double quote, but contained no whitespace,
3749 (b) If the value contained whitespace other than a space character (i.e.
3750 tabs or newlines or carriage returns) it was not quoted.
3751 This fix has been applied to the mysql and pgsql lookups by writing a
3752 separate quoting function and calling it from the lookup code. The fix
3753 should probably also be applied to nisplus, ibase and oracle lookups, but
3754 since I cannot test any of those, I have not disturbed their existing code.
3756 2. A hit in the callout cache for a specific address caused a log line with no
3757 reason for rejecting RCPT. Now it says "Previous (cached) callout
3758 verification failure".
3760 3. There was an off-by-one bug in the queryprogram router. An over-long
3761 return line was truncated at 256 instead of 255 characters, thereby
3762 overflowing its buffer with the terminating zero. As well as fixing this, I
3763 have increased the buffer size to 1024 (and made a note to document this).
3765 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3766 when Exim is waiting for an SMTP response from a remote server, Exim
3767 restarts its select() call on the socket, thereby resetting its timeout.
3768 This is not a problem when such interrupts are rare. Somebody set up a cron
3769 job to run exiwhat every 2 minutes, which is less than the normal select()
3770 timeout (5 or 10 minutes). This meant that the select() timeout never
3771 kicked in because it was always reset. I have fixed this by comparing the
3772 time when an interrupt arrives with the time at the start of the first call
3773 to select(). If more time than the timeout has elapsed, the interrupt is
3774 treated as a timeout.
3776 5. Some internal re-factoring in preparation for the addition of Sieve
3777 extensions (by MH). In particular, the "personal" test is moved to a
3778 separate function, and given an option for scanning Cc: and Bcc: (which is
3779 not set for Exim filters).
3781 6. When Exim created an email address using the login of the caller as the
3782 local part (e.g. when creating a From: or Sender: header line), it was not
3783 quoting the local part when it contained special characters such as @.
3785 7. Installed new OpenBSD configuration files.
3787 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3788 try to make them clearer.
3790 9. Callout options, other than the timeout value, were being ignored when
3791 verifying sender addresses in header lines. For example, when using
3793 verify = header_sender/callout=no_cache
3795 the cache was (incorrectly) being used.
3797 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3798 only the Cygwin environment.
3800 11. When return_path_on_delivery was set as a log selector, if different remote
3801 addresses in the same message used different return paths and parallel
3802 remote delivery occurred, the wrong values would sometimes be logged.
3803 (Whenever a remote delivery process finished, the return path value from
3804 the most recently started remote delivery process was logged.)
3806 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3807 lines when AUTH and/or TLS are in use. This is the "received protocol"
3808 field. Exim used to use "asmtp" for authenticated SMTP, without any
3809 indication (in the protocol name) for TLS use. Now it follows the RFC and
3810 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3811 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3812 names appear in log lines as well as in Received: header lines.
3814 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3815 extensions, and comparison tests, and to fix some bugs.
3817 14. Changes to the "personal" filter test:
3819 (1) The test was buggy in that it was just doing the equivalent of
3820 "contains" tests on header lines. For example, if a user's address was
3821 anne@some.where, the "personal" test would incorrectly be true for
3823 To: susanne@some.where
3825 This test is now done by extracting each address from the header in turn,
3826 and checking the entire address. Other tests that are part of "personal"
3827 are now done using regular expressions (for example, to check local parts
3828 of addresses in From: header lines).
3830 (2) The list of non-personal local parts in From: addresses has been
3831 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3832 taken from the Sieve specification recommendations.
3834 (3) If the message contains any header line starting with "List-" it is
3835 treated as non-personal.
3837 (4) The test for "circular" in the Subject: header line has been removed
3838 because it now seems ill-conceived.
3840 15. Minor typos in src/EDITME comments corrected.
3842 16. Installed latest exipick from John Jetmore.
3844 17. If headers_add on a router specified a text string that was too long for
3845 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3846 of string_sprintf() is now avoided.
3848 18. $message_body_size was not set (it was always zero) when running the DATA
3849 ACL and the local_scan() function.
3851 19. For the "mail" command in an Exim filter, no default was being set for
3852 the once_repeat time, causing a random time value to be used if "once" was
3853 specified. (If the value happened to be <= 0, no repeat happened.) The
3854 default is now 0s, meaning "never repeat". The "vacation" command was OK
3855 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3856 (I found it when inspecting the code).
3858 20. There is now an overall timeout for performing a callout verification. It
3859 defaults to 4 times the callout timeout, which applies to individual SMTP
3860 commands during the callout. The overall timeout applies when there is more
3861 than one host that can be tried. The timeout is checked before trying the
3862 next host. This prevents very long delays if there are a large number of
3863 hosts and all are timing out (e.g. when the network connections are timing
3864 out). The value of the overall timeout can be changed by specifying an
3865 additional sub-option for "callout", called "maxwait". For example:
3867 verify = sender/callout=5s,maxwait=20s
3869 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3870 to the end before writing, but this should make it even safer).
3872 22. Exim was forgetting that it had advertised PIPELINING for the second and
3873 subsequent messages on an SMTP connection. It was also not resetting its
3874 memory on STARTTLS and an internal HELO.
3876 23. When Exim logs an SMTP synchronization error within a session, it now
3877 records whether PIPELINING has been advertised or not.
3879 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3880 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3881 rather than long int.
3883 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3885 26. Added the never_mail option to autoreply.
3891 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3892 crash if the getsockname() call failed; this can happen if a connection is
3893 closed very soon after it is established. The problem was simply in the
3894 order in which certain operations were done, causing Exim to try to write
3895 to the SMTP stream before it had set up the file descriptor. The bug has
3896 been fixed by making things happen in the correct order.
3902 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3903 before the connection was closed, thus losing the rejection response.
3905 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3906 some early Solaris releases, but causes trouble in current releases where
3907 socklen_t is defined.
3909 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3912 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3915 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3916 that releasing the top of it at the end releases what was used for sub-
3917 expansions (unless the block got too big). However, discard this block if
3918 the first thing is a variable or header, so that we can use its block when
3919 it is dynamic (useful for very large $message_headers, for example).
3921 6. Lookups now cache *every* query, not just the most recent. A new, separate
3922 store pool is used for this. It can be recovered when all lookup caches are
3923 flushed. Lookups now release memory at the end of their result strings.
3924 This has involved some general refactoring of the lookup sources.
3926 7. Some code has been added to the store_xxx() functions to reduce the amount
3927 of flapping under certain conditions.
3929 8. log_incoming_interface used to affect only the <= reception log lines. Now
3930 it causes the local interface and port to be added to several more SMTP log
3931 lines, for example "SMTP connection from", and rejection lines.
3933 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3935 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3937 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3938 could be overwritten at the end of the current message (or the start of a
3939 new message if it was set in a HELO ACL). The value is now preserved for
3940 the duration of the SMTP connection.
3942 12. If a transport had a headers_rewrite setting, and a matching header line
3943 contained an unqualified address, that address was qualified, even if it
3944 did not match any rewriting rules. The underlying bug was that the values
3945 of the flags that permit the existence of unqualified sender and recipient
3946 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3947 non-local messages, and by -bnq for local messages) were not being
3948 preserved with the message after it was received.
3950 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3951 "next input=" as part of the text comprising the host identity instead of
3952 the correct text. The code was using the same buffer for two different
3953 strings. However, depending on which order the printing function evaluated
3954 its arguments, the bug did not always show up. Under Linux, for example, my
3955 test suite worked just fine.
3957 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3958 doesn't work with some older versions of Perl. It has been changed to "my",
3959 which in any case is probably the better facility to use.
3961 15. A really picky compiler found some instances of statements for creating
3962 error messages that either had too many or two few arguments for the format
3965 16. The size of the buffer for calls to the DNS resolver has been increased
3966 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3967 for addresses that have a lot of PTR records. This alleviates a problem; it
3968 does not fully solve it.
3970 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3971 buffer now truncates the list and logs the incident, which is the same
3972 action as happens when Exim is looking up a host name and its aliases.
3973 Previously in this situation something unpredictable would happen;
3974 sometimes it was "internal error: store_reset failed".
3976 18. If a server dropped the connection unexpectedly when an Exim client was
3977 using GnuTLS and trying to read a response, the client delivery process
3978 crashed while trying to generate an error log message.
3980 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3981 string, for example:
3983 warn message = H1: something\nH2: something
3985 the text was added as a single header line from Exim's point of view
3986 though it ended up OK in the delivered message. However, searching for the
3987 second and subsequent header lines using $h_h2: did not work. This has been
3988 fixed. Similarly, if a system filter added multiple headers in this way,
3989 the routers could not see them.
3991 20. Expanded the error message when iplsearch is called with an invalid key to
3992 suggest using net-iplsearch in a host list.
3994 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3995 ACLs are no longer actually imposed (and a message to that effect is
3998 22. If a "gecos" field in a passwd entry contained escaped characters, in
3999 particular, if it contained a \" sequence, Exim got it wrong when building
4000 a From: or a Sender: header from that name. A second bug also caused
4001 incorrect handling when an unquoted " was present following a character
4002 that needed quoting.
4004 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4005 was not being matched caselessly.
4007 24. Arranged for all hyphens in the exim.8 source to be escaped with
4010 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4011 properly. Recipient callout cache records were still being keyed to include
4012 the sender, even when use_sender was set false. This led to far more
4013 callouts that were necessary. The sender is no longer included in the key
4014 when use_sender is false.
4016 26. Added "control = submission" modifier to ACLs.
4018 27. Added the ${base62d: operator to decode base 62 numbers.
4020 28. dnsdb lookups can now access SRV records.
4022 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4023 the configuration file.
4025 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4026 (-v) mode. This makes the output for a verbose queue run more intelligible.
4028 31. Added a use_postmaster feature to recipient callouts.
4030 32. Added the $body_zerocount variable, containing the number of binary zero
4031 bytes in the message body.
4033 33. The time of last modification of the "new" subdirectory is now used as the
4034 "mailbox time last read" when there is a quota error for a maildir
4037 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4039 35. Added +ignore_unknown as a special item in host lists.
4041 36. Code for decoding IPv6 addresses in host lists is now included, even if
4042 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4043 address was recognized as an IP address, but was then not correctly decoded
4044 into binary, causing unexpected and incorrect effects when compared with
4051 1. Very minor rewording of debugging text in manualroute to say "list of
4052 hosts" instead of "hostlist".
4054 2. If verify=header_syntax was set, and a header line with an unqualified
4055 address (no domain) and a large number of spaces between the end of the
4056 name and the colon was received, the reception process suffered a buffer
4057 overflow, and (when I tested it) crashed. This was caused by some obsolete
4058 code that should have been removed. The fix is to remove it!
4060 3. When running in the test harness, delay a bit after writing a bounce
4061 message to get a bit more predictability in the log output.
4063 4. Added a call to search_tidyup() just before forking a reception process. In
4064 theory, someone could use a lookup in the expansion of smtp_accept_max_
4065 per_host which, without the tidyup, could leave open a database connection.
4067 5. Added the variables $recipient_data and $sender_data which get set from a
4068 lookup success in an ACL "recipients" or "senders" condition, or a router
4069 "senders" option, similar to $domain_data and $local_part_data.
4071 6. Moved the writing of debug_print from before to after the "senders" test
4074 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4075 problems for message scanning, either using a data ACL, or using
4076 local_scan() because the Received: header was not generated till after they
4077 were called (in order to set the time as the time of reception completion).
4078 I have revised the way this works. The header is now generated after the
4079 body is received, but before the ACL or local_scan() are called. After they
4080 are run, the timestamp in the header is updated.
4086 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4087 before starting a queue runner without re-exec. This happened only when
4088 deliver_drop_privilege was set or when the Exim user was set to root. The
4089 effect of the bug was that timeouts during subsequent deliveries caused
4090 crashes instead of being properly handled. The handler is now left at its
4091 default (and expected) setting.
4093 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4094 message, again when deliver_drop_privilege is set or Exim is run as root.
4095 The bug described in (1) was not present in this case, but the tidying up
4096 of the other signals was missing. I have made the two cases consistent.
4098 3. The ignore_target_hosts setting on a manualroute router was being ignored
4099 for hosts that were looked up using the /MX notation.
4101 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4104 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4105 operated on the sender address. After changing the $sender_address to <>
4106 for the sender address verify, Exim was re-instated it as the original
4107 (before rewriting) address, but remembering that it had rewritten it, so it
4108 wasn't rewriting it again. This bug also had the effect of breaking the
4109 sender address verification caching when the sender address was rewritten.
4111 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4112 This has been changed so that if the ip literal address matches
4113 ignore_target_hosts, the router declines.
4115 7. Added expansion conditions match_domain, match_address, and match_local_
4116 part (NOT match_host).
4118 8. The placeholder for the Received: header didn't have a length field set.
4120 9. Added code to Exim itself and to exim_lock to test for a specific race
4121 condition that could lead to file corruption when using MBX delivery. The
4122 issue is with the lockfile that is created in /tmp. If this file is removed
4123 after a process has opened it but before that process has acquired a lock,
4124 there is the potential for a second process to recreate the file and also
4125 acquire a lock. This could lead to two Exim processes writing to the file
4126 at the same time. The added code performs the same test as UW imapd; it
4127 checks after acquiring the lock that its file descriptor still refers to
4128 the same named file.
4130 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4131 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4134 11. Added the smtp_active_hostname option. If used, this will typically be made
4135 to depend on the incoming interface address. Because $interface_address is
4136 not set up until the daemon has forked a reception process, error responses
4137 that can happen earlier (such as "too many connections") no longer contain
4140 12. If an expansion in a condition on a "warn" statement fails because a lookup
4141 defers, the "warn" statement is abandoned, and the next ACL statement is
4142 processed. Previously this caused the whole ACL to be aborted.
4144 13. Added the iplsearch lookup type.
4146 14. Added ident_timeout as a log selector.
4148 15. Added tls_certificate_verified as a log selector.
4150 16. Added a global option tls_require_ciphers (compare the smtp transport
4151 option of the same name). This controls incoming TLS connections.
4153 17. I finally figured out how to make tls_require_ciphers do a similar thing
4154 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4155 before starting the TLS session.
4157 18. Tabs are now shown as \t in -bP output.
4159 19. If the log selector return_path_on_delivery was set, Exim crashed when
4160 bouncing a message because it had too many Received: header lines.
4162 20. If two routers both had headers_remove settings, and the first one included
4163 a superfluous trailing colon, the final name in the first list and the
4164 first name in the second list were incorrectly joined into one item (with a
4165 colon in the middle).
4171 1. Added -C and -D options to the exinext utility, mainly to make it easier
4172 to include in the automated testing, but these could be helpful when
4173 multiple configurations are in use.
4175 2. The exinext utility was not formatting the output nicely when there was
4176 an alternate port involved in the retry record key, nor when there was a
4177 message id as well (for retries that were specific to a specific message
4178 and a specific host). It was also confused by IPv6 addresses, because of
4179 the additional colons they contain. I have fixed the IPv4 problem, and
4180 patched it up to do a reasonable job for IPv6.
4182 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4183 delivery, the log line now contains "pipelined" if PIPELINING was used.
4185 4. An SMTP transport process used to panic and die if the bind() call to set
4186 an explicit outgoing interface failed. This has been changed; it is now
4187 treated in the same way as a connect() failure.
4189 5. A reference to $sender_host_name in the part of a conditional expansion
4190 that was being skipped was still causing a DNS lookup. This no longer
4193 6. The def: expansion condition was not recognizing references to header lines
4194 that used bh_ and bheader_.
4196 7. Added the _cache feature to named lists.
4198 8. The code for checking quota_filecount in the appendfile transport was
4199 allowing one more file than it should have been.
4201 9. For compatibility with Sendmail, the command line option
4209 and sets the incoming protocol and host name (for trusted callers). The
4210 host name and its colon can be omitted when only the protocol is to be set.
4211 Note the Exim already has two private options, -pd and -ps, that refer to
4212 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4213 "s", but I don't think that's a major issue.
4215 10. A number of refactoring changes to the code, none of which should affect
4218 (a) The number of logging options was getting close to filling up the
4219 32-bit word that was used as a bit map. I have split them into two classes:
4220 those that are passed in the argument to log_write(), and those that are
4221 only ever tested independently outside of that function. These are now in
4222 separate 32-bit words, so there is plenty of room for expansion again.
4223 There is no change in the user interface or the logging behaviour.
4225 (b) When building, for example, log lines, the code previously used a
4226 macro that called string_cat() twice, in order to add two strings. This is
4227 not really sufficiently general. Furthermore, there was one instance where
4228 it was actually wrong because one of the argument was used twice, and in
4229 one call a function was used. (As it happened, calling the function twice
4230 did not affect the overall behaviour.) The macro has been replaced by a
4231 function that can join an arbitrary number of extra strings onto a growing
4234 (c) The code for expansion conditions now uses a table and a binary chop
4235 instead of a serial search (which was left over from when there were very
4236 few conditions). Also, it now recognizes conditions like "pam" even when
4237 the relevant support is not compiled in: a suitably worded error message is
4238 given if an attempt is made to use such a condition.
4240 11. Added ${time_interval:xxxxx}.
4242 12. A bug was causing one of the ddress fields not to be passed back correctly
4243 from remote delivery subprocesses. The field in question was not being
4244 subsequently used, so this caused to problems in practice.
4246 13. Added new log selectors queue_time and deliver_time.
4248 14. Might have fixed a bug in maildirsizefile handling that threw up
4249 "unexpected character" debug warnings, and recalculated the data
4250 unnecessarily. In any case, I expanded the warning message to give more
4253 15. Added the message "Restricted characters in address" to the statements in
4254 the default ACL that block characters like @ and % in local parts.
4256 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4257 Three changes have been made:
4259 (a) There was a serious bug; a negative response to MAIL caused the whole
4260 recipient domain to be cached as invalid, thereby blocking all messages
4261 to all local parts at the same domain, from all senders. This bug has
4262 been fixed. The domain is no longer cached after a negative response to
4263 MAIL if the sender used is not empty.
4265 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4268 (c) A new callout option, "use_sender" has been added for people who want
4269 the modified behaviour.
4275 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4278 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4279 indeed breaks things for older releases.
4281 3. Added additional logging to the case where there is a problem reading data
4282 from a filter that is running in a subprocess using a pipe, in order to
4283 try to track down a specific problem.
4285 4. Testing facility fudge: when running in the test harness and attempting
4286 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4287 getting "No route to host". Convert this to a timeout.
4289 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4292 6. Some OS don't have socklen_t but use size_t instead. This affects the
4293 fifth argument of getsockopt() amongst other things. This is now
4294 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4295 can be set for individual OS. I have set it for SunOS5, OSF1, and
4296 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4297 some earlier ones do not.
4299 7. Change 4.30/15 was not doing the test caselessly.
4301 8. The standard form for an IPv6 address literal was being rejected by address
4302 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4303 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4304 this, as well as the form without the "IPv6" on the front (but only when
4305 address literals are enabled, of course).
4307 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4309 10. Exim crashed if a message with an empty sender address specified by -f
4310 encountered a router with an errors_to setting. This could be provoked only
4311 by a command such as
4315 where an empty string was supplied; "<>" did not hit this bug.
4317 11. Installed PCRE release 4.5.
4319 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4320 remained set. It is now erased.
4322 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4323 times from message ids (which are base 36 rather than the normal 62).
4325 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4326 were being counted as actual protocol errors, and logged if the log
4327 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4328 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4329 and DATA following a set of rejected RCPTs do not count as protocol errors.
4330 In other words, Exim assumes they were pipelined, though this may not
4331 actually be the case. Of course, in all cases the client gets an
4332 appropriate error code.
4334 15. If a lookup fails in an ACL condition, a message about the failure may
4335 be available; it is used if testing the ACL cannot continue, because most
4336 such messages specify what the cause of the deferral is. However, some
4337 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4338 that caused an old message to be retained and used if a later statement
4339 caused a defer, replacing the real cause of the deferral.
4341 16. If an IP address had so many PTR records that the DNS lookup buffer
4342 was not large enough to hold them, Exim could crash while trying to process
4343 the truncated data. It now detects and logs this case.
4345 17. Further to 4.21/58, another change has been made: if (and only if) the
4346 first line of a message (the first header line) ends with CRLF, a bare LF
4347 in a subsequent header line has a space inserted after it, so as not to
4348 terminate the header.
4350 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4351 unnecessarily, used atoi() instead of strtol(), and didn't check the
4352 termination when getting file sizes from file names by regex.
4354 19. Completely re-implemented the support for maildirsize files, in the light
4355 of a number of problems with the previous contributed implementation
4356 (4.30/29). In particular:
4358 . If the quota is zero, the maildirsize file is maintained, but no quota is
4361 . If the maildir directory does not exist, it is created before any attempt
4362 to write a maildirsize file.
4364 . The quota value in the file is just a cache; if the quota is changed in
4365 the transport, the new value overrides.
4367 . A regular expression is available for excluding directories from the
4370 20. The autoreply transport checks the characters in options that define the
4371 message's headers; it allows continued headers, but it was checking with
4372 isspace() after an embedded newline instead of explicitly looking for a
4375 21. If all the "regular" hosts to which an address was routed had passed their
4376 expiry times, and had not reached their retry times, the address was
4377 bounced, even if fallback hosts were defined. Now Exim should go on to try
4380 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4381 equivalent code in the SMTP transport. Some hosts send humungous responses
4382 to HELO/EHLO, more than 1024 it seems.
4384 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4385 gives compiler warnings in some environments. I've now done it "properly",
4388 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4389 (because of problems with the built-in one) was declared to return uschar *
4390 instead of char *, causing compiler failure.
4392 25. Fixed a file descriptor leak when processing alias/forward files.
4394 26. Fixed a minor format string issue in dbfn.c.
4396 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4398 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4399 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4402 29. When a sender address is verified, it is cached, to save repeating the test
4403 when there is more than one recipient in a message. However, when the
4404 verification involves a callout, it is possible for different callout
4405 options to be set for different recipients. It is too complicated to keep
4406 track of this in the cache, so now Exim always runs a verification when a
4407 callout is required, relying on the callout cache for the optimization.
4408 The overhead is duplication of the address routing, but this should not be
4411 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4412 to be verified with callout=postmaster, and the main callout worked but the
4413 postmaster check failed, the verification correctly failed. However, if a
4414 subsequent RCPT command asked for sender verification *without* the
4415 postmaster check, incorrect caching caused this verification also to fail,
4418 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4419 it was not caching the DNS options (qualify_single, search_parents) that
4420 were used when the lookup failed. A subsequent lookup with different
4421 options therefore always gave the same answer, though there were cases
4422 where it should not have. (Example: a "domains = !$mx_any" option on a
4423 dnslookup router: the "domains" option is always processed without any
4424 widening, but the router might have qualify_single set.) Now Exim uses the
4425 cached value only when the same options are set.
4427 32. Added John Jetmore's "exipick" utility to the distribution.
4429 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4430 than a timeout (e.g. a certificate is required, and is not provided), an
4431 Exim server now closes the connection immediately. Previously it waited for
4432 the client to close - but if the client is SSL, it seems that they each
4433 wait for each other, leading to a delay before one of them times out.
4435 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4436 maintained 0.8.x compatibility because I don't think many are using it, and
4437 it is clearly obsolete.
4439 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4442 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4443 client certificate was expired. A simple patch fixes this, though I don't
4444 understand the full logic of why the verify callback is called multiple
4447 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4448 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4449 which causes problems with some clients (such as the Certicom SSL Plus
4450 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4451 disables the coutermeasure allowing Eudora to connect."
4453 38. Exim was not checking that a write() to a log file succeeded. This could
4454 lead to Bad Things if a log got too big, in particular if it hit a file
4455 size limit. Exim now panics and dies if it cannot write to a log file, just
4456 as it does if it cannot open a log file.
4458 39. Modified OS/Makefile-Linux so that it now contains
4460 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4462 The two -D definitions ensure that Exim is compiled with large file
4463 support, which makes it possible to handle log files that are bigger than
4466 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4467 instance) a domain was checked against a named list that involved a lookup,
4468 causing $domain_data to be set, then another domain was checked against the
4469 same list, then the first domain was re-checked, the value of $domain_data
4470 after the final check could be wrong. In particular, if the second check
4471 failed, it could be set empty. This bug probably also applied to
4474 41. The strip_trailing_dot option was not being applied to the address given
4475 with the -f command-line option.
4477 42. The code for reading a message's header from the spool was incrementing
4478 $received_count, but never initializing it. This meant that the value was
4479 incorrect (doubled) while delivering a message in the same process in which
4480 it was received. In the most common configuration of Exim, this never
4481 happens - a fresh exec is done - but it can happen when
4482 deliver_drop_privilege is set.
4484 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4485 it now includes up to 150 characters of the unexpected data in the log
4488 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4489 and building data strings. The size of both of these buffers was 10 000
4490 bytes - far larger than anybody would *ever* want, thought I. Needless to
4491 say, somebody hit the limit. I have increased the maximum line length to
4492 20 000 and the maximum data length of concatenated lines to 100 000. I have
4493 also fixed two bugs, because there was no checking on these buffers. Tsk,
4494 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4495 buffer is too small.
4497 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4498 lsearch lookups. Now it does.
4500 46. When parsing a route_list item in a manualroute router, a fixed-length
4501 buffer was used for the list of hosts. I made this 1024 bytes long,
4502 thinking that nobody would ever have a list of hosts that long. Wrong.
4503 Somebody had a whole pile of complicated expansion conditions, and the
4504 string was silently truncated, leading to an expansion error. It turns out
4505 that it is easier to change to an unlimited length (owing to other changes
4506 that have happened since this code was originally written) than to build
4507 structure for giving a limitation error. The length of the item that
4508 expands into the list of hosts is now unlimited.
4510 47. The lsearch lookup could not handle data where the length of text line was
4511 more than 4095 characters. Such lines were truncated, leading to shortened
4512 data being returned. It should now handle lines of any length.
4514 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4515 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4518 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4519 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4520 no longer happens in comment lines. A list of replacements is now placed
4521 at the head of all of the source files, except those whose only change is
4522 to replace PERL_COMMAND in the very first #! line.
4524 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4525 messages on the queue, with a bottom-up merge sort, using code contributed
4526 by Michael Haardt. This should make operations like -bp somewhat faster on
4527 large queues. It won't affect queue runners, except when queue_run_in_order
4530 51. Installed eximstats 1.31 in the distribution.
4532 52. Added support for SRV lookups to the dnslookup router.
4534 53. If an ACL referred to $message_body or $message_body_end, the value was not
4535 reset for any messages that followed in the same SMTP session.
4537 54. The store-handling optimization for building very long strings was not
4538 differentiating between the different store pools. I don't think this
4539 actually made any difference in practice, but I've tidied it.
4541 55. While running the routers to verify a sender address, $sender_address
4542 was still set to the sender address. This is wrong, because when routing to
4543 send a bounce to the sender, it would be empty. Therefore, I have changed
4544 it so that, while verifying a sender address, $sender_address is set to <>.
4545 (There is no change to what happens when verifying a recipient address.)
4547 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4548 target A or AAAA records (if not already returned) without resetting the
4549 qualify_single or search_parents options of the DNS resolver. These are
4550 inappropriate in this case because the targets of MX and SRV records must
4551 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4552 target that, when qualified, matched something in the local domain. These
4553 two options are now turned off when doing these lookups.
4555 57. It seems that at least some releases of Reiserfs (which does not have the
4556 concept of a fixed number of inodes) returns zero and not -1 for the
4557 number of available inodes. This interacted badly with check_spool_inodes,
4558 which assumed that -1 was the "no such thing" setting. What I have done is
4559 to check that the total number of inodes is greater than zero before doing
4560 the test of how many are available.
4562 58. When a "warn" ACL statement has a log_message modifier, the message is
4563 remembered, and not repeated. This is to avoid a lot of repetition when a
4564 message has many recipients that cause the same warning to be written.
4565 Howewer, Exim was preserving the list of already written lines for an
4566 entire SMTP session, which doesn't seem right. The memory is now reset if a
4567 new message is started.
4569 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4570 debugging output unless log_rewrite was also set.
4572 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4573 of (int)(handle) when we know that handle contains (void *)(-1).
4575 61. The Exim daemon panic-logs an error return when it closes the incoming
4576 connection. However "connection reset by peer" seems to be common, and
4577 isn't really an error worthy of noting specially, so that particular error
4580 62. When Exim is trying to find all the local interfaces, it used to panic and
4581 die if the ioctl to get the interface flags failed. However, it seems that
4582 on at least one OS (Solaris 9) it is possible to have an interface that is
4583 included in the list of interfaces, but for which you get a failure error
4584 for this call. This happens when the interface is not "plumbed" into a
4585 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4586 failure of the "get flags" call assumes that the interface is down.
4588 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4589 makes life easier for people who are doing arithmetic on fields extracted
4590 from dates, where you often get leading zeros that should not be
4591 interpreted as octal.
4593 64. Added qualify_domain to the redirect router, to override the global
4596 65. If a pathologically long header line contained very many addresses (the
4597 report of this problem mentioned 10 000) and each of them was rewritten,
4598 Exim could use up a very large amount of memory. (It kept on making new
4599 copies of the header line as it rewrote, and never released the old ones.)
4600 At the expense of a bit more processing, the header rewriting function has
4601 been changed so that it no longer eats memory in this way.
4603 66. The generation of the Received: header has been moved from the time that a
4604 message starts to be received, to the time that it finishes. The timestamp
4605 in the Received: header should now be very close to that of the <= log
4606 line. There are two side-effects of this change:
4608 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4609 logged header lines no longer include the local Received: line, because
4610 it has not yet been created. The same applies to a copy of the message
4611 that is returned to a non-SMTP sender when a message is rejected.
4613 (b) When a filter file is tested using -bf, no additional Received: header
4614 is added to the test message. After some thought, I decided that this
4617 This change does not affect the value of $received_for. It is still set
4618 after address rewriting, but before local_scan() is called.
4620 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4622 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4623 gave an unhelpful panic error message, and a defer error. I have managed to
4624 change this behaviour so that it now rejects any supplied certificate,
4625 which seems right, as the list of acceptable certificates is empty.
4627 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4628 gave an unhelpful defer error. I have not managed to make this reject any
4629 supplied certificates, but the error message it gives is "no certificate
4630 supplied", which is not helpful.
4632 70. exigrep's output now also includes lines that are not associated with any
4633 message, but which match the given pattern. Implemented by a patch from
4634 Martin Sluka, which also tidied up the Perl a bit.
4636 71. Recipient callout verification, like sender verification, was using <> in
4637 the MAIL FROM command. This isn't really the right thing, since the actual
4638 sender may affect whether the remote host accepts the recipient or not. I
4639 have changed it to use the actual sender in the callout; this means that
4640 the cache record is now keyed on a recipient/sender pair, not just the
4641 recipient address. There doesn't seem to be a real danger of callout loops,
4642 since a callout by the remote host to check the sender would use <>.
4643 [SEE ABOVE: changed after hitting problems.]
4645 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4646 temporary errors. However, in the case of such a code being given after
4647 the end of a data transmission (i.e. after ".") Exim was failing to write
4648 a retry record for the message. (Yes, there was some broken host that was
4649 actually sending 8xx at this point.)
4651 73. An unknown lookup type in a host list could cause Exim to panic-die when
4652 the list was checked. (An example that provoked this was putting <; in the
4653 middle of a list instead of at the start.) If this happened during a DATA
4654 ACL check, a -D file could be left lying around. This kind of configuration
4655 error no longer causes Exim to die; instead it causes a defer errror. The
4656 incident is still logged to the main and panic logs.
4658 74. Buglet left over from Exim 3 conversion. The message "too many messages
4659 in one connection" was written to the rejectlog but not the mainlog, except
4660 when address rewriting (yes!) was being logged.
4662 75. Added write_rejectlog option.
4664 76. When a system filter was run not as root (that is, when system_filter_user
4665 was set), the values of the $n variables were not being returned to the
4666 main process; thus, they were not subsequently available in the $sn
4669 77. Added +return_path_on_delivery log selector.
4671 78. A connection timeout was being treated differently from recipients deferred
4672 when testing hosts_max_try with a message that was older than the host's
4673 retry timeout. (The host should not be counted, thus allowing all hosts to
4674 be tried at least once before bouncing.) This may have been the cause of an
4675 occasionally reported bug whereby a message would remain on the queue
4676 longer than the retry timeout, but would be bounced if a delivery was
4677 forced. I say "may" because I never totally pinned down the problem;
4678 setting up timeout/retry tests is difficult. See also the next item.
4680 79. The ultimate address timeout was not being applied to errors that involved
4681 a combination of host plus message (for example, a timeout on a MAIL
4682 command). When an address resolved to a number of possible hosts, and they
4683 were not all tried for each delivery (e.g. because of hosts_max_try), a
4684 message could remain on the queue longer than the retry timeout.
4686 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4689 81. Fixed an obscure SMTP outgoing bug which required at least the following
4690 conditions: (a) there was another message waiting for the same server;
4691 (b) the server returned 5xx to all RCPT commands in the first message so
4692 that the message was not completed; (c) the server dropped the connection
4693 or gave a negative response to the RSET that Exim sends to abort the
4694 transaction. The observed case was a dropped connection after DATA that had
4695 been sent in pipelining mode. That is, the server had advertised PIPELINING
4696 but was not implementing it correctly. The effect of the bug was incorrect
4697 behaviour, such as trying another host, and this could lead to a crash.
4703 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4704 and daemon.c were passed as pointers to ints; they should have been
4705 pointers to socklen_t variables (which are typically unsigned ints).
4707 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4710 3. Fixed a really odd bug that affected only the testing scheme; patching a
4711 certain fixed string in the binary changed the value of another string that
4712 happened to be identical to the end of the original first string.
4714 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4715 name", it returns that address as the IP address. On some operating
4716 systems (e.g. Solaris), it also passes back the IP address string as the
4717 "host name". However, on others (e.g. Linux), it passes back an empty
4718 string. Exim wasn't checking for this, and was changing the host name to an
4719 empty string, assuming it had been canonicized.
4721 5. Although rare, it is permitted to have more than one PTR record for a given
4722 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4723 all the names associated with an address, because they do in Solaris.
4724 However, it seems that they do not in Linux for data that comes from the
4725 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4726 I found this out when I moved to a new Linux workstation and tried to run
4727 the Exim test suite.
4729 To get round this problem I have changed the code so that it now does its
4730 own call to the DNS to look up PTR records when searching for a host name.
4731 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4732 addresses that are only in /etc/hosts are still found.
4734 This behaviour is, however, controlled by an option called host_lookup_
4735 order, which defaults to "bydns:byaddr". If people want to use the other
4736 order, or indeed, just use one or the other means of lookup, they can
4737 specify it in this variable.
4739 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4740 some operating systems, this comes back from gethostbyaddr() as an empty
4741 string, and this is what Exim used to test for. However, it seems that in
4742 other systems, "." is yielded. Exim now tests for this case too.
4744 7. The values of check_spool_space and check_log_space are now held internally
4745 as a number of kilobytes instead of an absolute number of bytes. If a
4746 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4747 kilobyte. This means that much larger values can be stored.
4749 8. Exim monitor: an attempt to get the action menu when not actually pointing
4750 at a message produces an empty menu entitled "No message selected". This
4751 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4752 no entries in it ("Shell widget menu has zero width and/or height"). So I
4753 have added a single, blank menu entry in this case.
4755 9. Added ${quote_local_part.
4757 10. MIME decoding is now applied to the contents of Subject: header lines when
4760 11. Now that a reference to $sender_host_address automatically causes a reverse
4761 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4762 host lookup before query-style lookups in lists that might use this
4763 variable. This has therefore been abolished, and the "net-" prefix is no
4764 longer necessary for query-style lookups.
4766 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4767 have been a typo for LFLAGS, so it has been changed.
4769 13. The install script calls Exim with "-C /dev/null" in order to find the
4770 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4771 to be output. Howeve, since Exim outputs its version number before the
4772 error, it didn't break the script. It just looked ugly. I fixed this by
4773 always allowing "-C /dev/null" if the caller is root.
4775 14. Ignore overlarge ACL variable number when reading spool file - insurance
4776 against a later release with more variables having written the file.
4778 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4779 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4780 this, as well as the form without the "IPv6" on the front.
4782 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4783 OS/Makefile-Darwin file.
4785 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4786 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4788 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4789 variable. However, some OS use ints for inodes. I've added cast to long int
4790 to get rid of the compiler warning.
4792 19. I had forgotten to lock out "/../" in configuration file names when
4793 ALT_CONFIG_PREFIX was set.
4795 20. Routers used for verification do not need to specify transports. However,
4796 if such a router generated a host list, and callout was configured, Exim
4797 crashed, because it could not find a port number from the (non-existent)
4798 transport. It now assumes port 25 in this circumstance.
4800 21. Added the -t option to exigrep.
4802 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4803 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4804 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4805 them is set, LOOKUP_LSEARCH is forced.
4807 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4809 24. Added sender and host information to the "rejected by local_scan()" log
4810 line; previously there was no indication of these.
4812 25. Added .include_if_exists.
4814 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4815 Linux. It turns out that not all file systems support this. Apparently some
4816 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4817 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4818 not supported on the file descriptor, is now ignored when Exim is trying to
4819 sync a directory. This applies only to Linux.
4821 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4823 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4824 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4827 29. Added support for maildirsize files from supplied patch (modified a bit).
4829 30. The use of :fail: followed by an empty string could lead Exim to respond to
4830 sender verification failures with (e.g.):
4832 550 Verification failed for <xxx>
4833 550 Sender verify failed
4835 where the first response line was missing the '-' that indicates it is not
4836 the final line of the response.
4838 31. The loop for finding the name of the user that called Exim had a hardwired
4839 limit of 10; it now uses the value of finduser_retries, which is used for
4840 all other user lookups.
4842 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4845 33. Exim was neglecting to zero errno before one call of strtol() when
4846 expanding a string and expecting an integer value. On some systems this
4847 resulted in spurious "integer overflow" errors. Also, it was casting the
4848 result into an int without checking.
4850 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4851 did not work. The code looks as if it has *never* worked, though it appears
4852 to have been documented since at least releast 1.62. I have made it work.
4854 35. The "timeout_DNS" error in retry rules, also documented since at least
4855 1.62, also never worked. As it isn't clear exactly what this means, and
4856 clearly it isn't a major issue, I have abolished the feature by treating it
4857 as "timeout", and writing a warning to the main and panic logs.
4859 36. The display of retry rules for -brt wasn't always showing the error code
4862 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4863 timeout_connect_A, timeout_connect_MX.
4865 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4866 to the empty sender.
4868 39. The daemon was not analysing the content of -oX till after it had closed
4869 stderr and disconnected from the controlling terminal. This meant that any
4870 syntax errors were only noted on the panic log, and the return code from
4871 the command was 0. By re-arranging the code a little, I've made the
4872 decoding happen first, so such errors now appear on stderr, and the return
4873 code is 1. However, the actual setting up of the sockets still happens in
4874 the disconnected process, so errors there are still only recorded on the
4877 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4878 connections (as happens on some IP stacks) was logged at start up time as
4879 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4880 it from "IPv6 and IPv4", which means that two separate sockets are being
4883 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4884 says whether AF_INET or AF_INET6 was passed as an argument.
4886 42. Exiwhat output was messed up when time zones were included in log
4889 43. Exiwhat now gives more information about the daemon's listening ports,
4890 and whether -tls-on-connect was used.
4892 44. The "port" option of the smtp transport is now expanded.
4894 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4895 silently ignored. Now an error message is written to the main and panic
4898 46. There's a new ACL modifier called "logwrite" which writes to a log file
4899 as soon as it is encountered.
4901 47. Added $local_user_uid and $local_user_gid at routing time.
4903 48. Exim crashed when trying to verify a sender address that was being
4906 49. Exim was recognizing only a space character after ".include". It now also
4907 recognizes a tab character.
4909 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4910 extracting the relevant information from the specification. The man page no
4911 longer contains scrambled data for the -d option, and I've added a section
4912 at the front about calling Exim under different names.
4914 51. Added "extra_headers" argument to the "mail" command in filter files.
4916 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4919 53. Installed eximstats 1.29.
4921 54. Added transport_filter_timeout as a generic transport option.
4923 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4924 Cc: header lines. This was required by RFC 822, but it not required by RFC
4927 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4928 incoming messages that did not have them. Now it does so only if the
4929 message originates locally, that is, if there is no associated remote host
4930 address. When Resent- header lines are present, this applies to the Resent-
4931 lines rather than the non-Resent- lines.
4933 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4934 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4936 58. Messages for configuration errors now include the name of the main
4937 configuration file - useful now that there may be more than one file in a
4938 list (.included file names were always shown).
4940 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4941 for those rare installations that do not start the daemon as root or run it
4942 setuid root. I've cut out the call to initgroups() if the daemon is not
4945 60. The Exim user and group can now be bound into the binary as text strings
4946 that are looked up at the start of Exim's processing.
4948 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4950 62. Added $mailstore_basename variable.
4952 63. Installed patch to sieve.c from Michael Haardt.
4954 64. When Exim failed to open the panic log after failing to open the main log,
4955 the original message it was trying to log was written to stderr and debug
4956 output, but if they were not available (the usual case in production), it
4957 was lost. Now it is written to syslog before the two lines that record the
4958 failures to open the logs.
4960 65. Users' Exim filters run in subprocesses under the user's uid. It is
4961 possible for a "deliver" command or an alias in a "personal" command to
4962 provoke an address rewrite. If logging of address rewriting is configured,
4963 this fails because the process is not running as root or exim. There may be
4964 a better way of dealing with this, but for the moment (because 4.30 needs
4965 to be released), I have disabled address rewrite logging when running a
4966 filter in a non-root, non-exim process.
4972 1. The buildconfig auxiliary program wasn't quoting the value set for
4973 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4974 not defined. This bug was masked in 4.22 by the effect that was fixed in
4977 2. Some messages that were rejected after a message id was allocated were
4978 shown as "incomplete" by exigrep. It no longer does this for messages that
4979 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4981 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4982 have allow_domain_literals set, the ID did not get logged in the <= line.
4983 Domain literals are now always recognized in Message-ID: header lines.
4985 4. The first argument for a ${extract expansion item is the key name or field
4986 number. Leading and trailing spaces in this item were not being ignored,
4987 causing some misleading effects.
4989 5. When deliver_drop_privilege was set, single queue runner processes started
4990 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4991 same command in the process it spins off) were not dropping privilege.
4993 6. When the daemon running as "exim" started a queue runner, it always
4994 re-executed Exim in the spun-off process. This is a waste of effort when
4995 deliver_drop_privilege is set. The new process now just calls the
4996 queue-runner function directly.
5002 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5005 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5006 ignored. Though the use of -oP was forcing the writing of a pid file, it
5007 was always written to the default place.
5009 3. If the message "no IP address found for host xxxx" is generated during
5010 incoming verification, it is now followed by identification of the incoming
5011 connection (so you can more easily find what provoked it).
5013 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5015 5. Added some features to "harden" Exim a bit more against certain attacks:
5017 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5018 be put in Local/Makefile. This is like the never_users runtime option,
5019 but it cannot be overridden. The default setting is "root".
5021 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5022 prefix string with which any file named in a -C command line option
5025 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5026 is retained for -C and -D only if the caller of Exim is root. Without
5027 it, the exim user may also use -C and -D and retain privilege.
5029 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5030 command line option is disabled.
5032 6. Macro names set by the -D option must start with an upper case letter, just
5033 like macro names defined in the configuration file.
5035 7. Added "dereference=" facility to LDAP.
5037 8. Two instances of the typo "uknown" in the source files are fixed.
5039 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5040 the Configure-Makefile script screwed up while processing it.
5042 10. Incorporated PCRE 4.4.
5044 11. The SMTP synchronization check was not operating right at the start of an
5045 SMTP session. For example, it could not catch a HELO sent before the client
5046 waited for the greeting. There is now a check for outstanding input at the
5047 point when the greeting is written. Because of the duplex, asynchronous
5048 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5049 way, but not yet received, when the check is performed.
5051 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5052 on TCP/IP sockets, because this apparently causes some broken clients to
5055 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5056 unchanged) from the Cygwin maintainer.
5058 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5059 was supported instead of testing for mbx. Effectively a typo.
5061 15. The spa authenticator server code was not checking that the input it
5062 received was valid base64.
5064 16. The debug output line for the "set" modifier in ACLs was not showing the
5065 name of the variable that was being set.
5067 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5069 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5070 lookup on its own. Something else was needed to trigger the lookup. For
5071 example, a match in host_lookup or the need for a host name in a host list.
5072 Now, if $sender_host_name is referenced and the host name has not yet been
5073 looked up, a lookup is performed. If the lookup fails, the variable remains
5074 empty, and $host_lookup_failed is set to "1".
5076 19. Added "eqi" as a case-independent comparison operator.
5078 20. The saslauthd authentication condition could segfault if neither service
5079 nor realm was specified.
5081 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5082 error message that was logged was misleading, and incoming SMTP
5083 connections were dropped. The message is now more accurate, and temporary
5084 errors are given to SMTP connections.
5086 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5087 (except RSET) with a 421 error, until QUIT is received. However, it was
5088 failing to send a response to QUIT.
5090 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5091 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5092 ACL is now run after the helo_try_verify_hosts code.
5094 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5095 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5096 case-independent, so other case variants are also recognized). Apparently
5097 some systems use these upper case variants.
5099 25. If more than two messages were waiting for the same host, and a transport
5100 filter was specified for the transport, Exim sent two messages over the
5101 same TCP/IP connection, and then failed with "socket operation on non-
5102 socket" when it tried to send the third.
5104 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5106 27. The extern definition of crypt16() in expand.c was not being excluded when
5107 the OS had its own crypt16() function.
5109 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5110 as a preferred synonym for return_size_limit, both as an option and as an
5113 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5115 30. Changed the default configuration ACL to relax the local part checking rule
5116 for addresses that are not in any local domains. For these addresses,
5117 slashes and pipe symbols are allowed within local parts, but the sequence
5118 /../ is explicitly forbidden.
5120 31. SPA server authentication was not clearing the challenge buffer before
5123 32. log_message in a "warn" ACL statement was writing to the reject log as
5124 well as to the main log, which contradicts the documentation and doesn't
5125 seem right (because no rejection is happening). So I have stopped it.
5127 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5128 However, I am unable to do any testing of this.
5130 34. Fixed an infelicity in the appendfile transport. When checking directories
5131 for a mailbox, to see if any needed to be created, it was accidentally
5132 using path names with one or more superfluous leading slashes; tracing
5133 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5135 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5136 contents are added to the log line that is written for every discarded
5137 recipient. (Previously a log_message setting was ignored.)
5139 36. The ${quote: operator now quotes the string if it is empty.
5141 37. The install script runs exim in order to find its version number. If for
5142 some reason other than non-existence or emptiness, which it checks, it
5143 could not run './exim', it was installing it with an empty version number,
5144 i.e. as "exim-". This error state is now caught, and the installation is
5147 38. An argument was missing from the function that creates an error message
5148 when Exim fails to connect to the socket for saslauthd authentication.
5149 This could cause Exim to crash, or give a corrupted message.
5151 39. Added isip, isip4, and isip6 to ${if conditions.
5153 40. The ACL variables $acl_xx are now saved with the message, and can be
5154 accessed later in routers, transports, and filters.
5156 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5157 strings in the file are not string-expanded.
5159 42. If a MAIL command specified a SIZE value that was too large to fit into an
5160 int variable, the check against message_size_limit failed. Such values are
5161 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5162 day this will have to be increased, but I don't think I want to be around
5163 when emails are that large.
5170 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5171 iconv() is not standard in FreeBSD.
5173 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5174 IPv6 enabled. The observed symptom was a segmentation fault on return from
5175 the function os_common_find_running_interfaces() in src/os.c.
5177 3. In the check_special_case() function in daemon.c I had used "errno" as an
5178 argument name, which causes warnings on some systems. This was basically a
5179 typo, since it was named "eno" in the comments!
5181 4. The code that waits for the clock to tick (at a resolution of some fraction
5182 of a second) so as to ensure message-id uniqueness was always waiting for
5183 at least one whole tick, when it could have waited for less. [This is
5184 almost certainly not relevant at current processor speeds, where it is
5185 unlikely to ever wait at all. But we try to future-proof.]
5187 5. The function that sleeps for a time interval that includes fractions of a
5188 second contained a race. It did not block SIGALRM between setting the
5189 timer, and suspending (a couple of lines later). If the interval was short
5190 and the sigsuspend() was delayed until after it had expired, the suspension
5191 never ended. On busy systems this could lead to processes getting stuck for
5194 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5195 process, before it forks any delivery processes. The open lookup caching
5196 mechanism meant that the open file or database connection was passed into
5197 the delivery process. The problem was that delivery processes always tidy
5198 up cached lookup data. This could cause a problem for the next delivery
5199 process started by the queue runner, because the external queue runner
5200 process does not know about the closure. So the next delivery process
5201 still has data in the lookup cache. In the case of a file lookup, there was
5202 no problem because closing a file descriptor in a subprocess doesn't affect
5203 the parent. However, if the lookup was caching a connection to a database,
5204 the connection was closed, and the second delivery process was likely to
5205 see errors such as "PGSQL: query failed: server closed the connection
5206 unexpectedly". The problem has been fixed by closing all cached lookups
5207 in a queue runner before running a delivery process.
5209 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5210 seem to have the "const" qualifier which it has on other OS. I've
5213 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5214 *of the same type* (client or server) with the same public name that an
5215 error should be diagnosed.
5217 9. When Exim looked up a host name for an IP address, but failed to find the
5218 original IP address when looking up the host name (a safety check), it
5219 output the message "<ip address> does not match any IP for NULL", which was
5220 confusing, to say the least. The bug was that the host name should have
5221 appeared instead of "NULL".
5223 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5224 user that is built into the binary, and the -C or -D options is used, root
5225 privilege is dropped before the configuration file is read. In addition,
5226 logging is switched to stderr instead of the normal log files. If the
5227 configuration then re-defines the Exim user, the unprivileged environment
5228 is probably not what is expected, so Exim logs a panic warning message (but
5231 However, if deliver_drop_privilege is set, the unprivileged state may well
5232 be exactly what is intended, so the warning has been cut out in that case,
5233 and Exim is allowed to try to write to its normal log files.
5239 1. smtp_return_error_details was not giving details for temporary sender
5240 or receiver verification errors.
5242 2. Diagnose a configuration error if two authenticators have the same public
5245 3. Exim used not to create the message log file for a message until the first
5246 delivery attempt. This could be confusing when incoming messages were held
5247 for policy or load reasons. The message log file is now created at the time
5248 the message is received, and an initial "Received" line is written to it.
5250 4. The automatically generated man page for command line options had a minor
5251 bug that caused no ill effects; however, a more serious problem was that
5252 the procedure for building the man page automatically didn't always
5253 operate. Consequently, release 4.20 contains an out-of-date version. This
5254 shouldn't happen again.
5256 5. When building Exim with embedded Perl support, the script that builds the
5257 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5258 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5260 6. The freeze_tell option was not being used for messages that were frozen on
5261 arrival, either by an ACL or by local_scan().
5263 7. Added the smtp_incomplete_transaction log selector.
5265 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5266 was accepting AUTH without a new EHLO.
5268 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5269 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5272 10. Logging of TCP/IP connections (when configured) now happens in the main
5273 daemon process instead of the child process, so that the TCP/IP connection
5274 count is more accurate (but it can never be perfect).
5276 11. The use of "drop" in a nested ACL was not being handled correctly in the
5277 outer ACL. Now, if condition failure induced by the nested "drop" causes
5278 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5279 or "require"), the connection is dropped.
5281 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5282 that yield "discard" can now be used with an "accept" or a "discard" verb,
5283 but an error is generated for any others (because I can't see a useful way
5284 to define what should happen).
5286 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5287 are now processed in the same way as lines in the Exim configuration file.
5288 In particular, continuation lines are supported.
5290 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5292 15. Added -ti meaning -t -i.
5294 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5295 domains, and warn by logging if others are found.
5297 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5298 returned from GIFCONF when Exim is trying to find the list of interfaces on
5299 a host. The code in os.c has been modified to copy each ifreq to an aligned
5300 structure in all cases.
5302 Also, in some cases, the returned ifreq's were being copied to a 'struct
5303 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5304 means the last couple of bytes of an IPv6 address could be chopped if the
5305 ifreq contained only a normal sockaddr (14 bytes storage).
5307 18. Named domain lists were not supported in the hosts_treat_as_local option.
5308 An entry such as +xxxx was not recognized, and was treated as a literal
5311 19. Ensure that header lines added by a DATA ACL are included in the reject log
5312 if the ACL subsequently rejects the message.
5314 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5315 MD5 (which is deprecated).
5317 21. When testing a filter file using -bf, Exim was writing a message when it
5318 took the sender from a "From " line in the message, but it was not doing so
5319 when it took $return_path from a Return-Path: header line. It now does.
5321 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5322 with a valid header line field name (a series of printing characters
5323 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5325 23. Changed "disc" in the source to "disk" to conform to the documentation and
5326 the book and for uniformity.
5328 24. Ignore Sendmail's -Ooption=value command line item.
5330 25. When execve() failed while trying to run a command in a pipe transport,
5331 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5332 could be confused with a return value of 69 from the command itself. This
5333 has been changed to 127, the value the shell returns if it is asked to run
5334 a non-existent command. The wording for the related log line suggests a
5335 non-existent command as the problem.
5337 26. If received_header_text expands to an empty string, do not add a Received:
5338 header line to the message. (Well, it adds a token one on the spool, but
5339 marks it "old" so that it doesn't get used or transmitted.)
5341 27. Installed eximstats 1.28 (addition of -nt option).
5343 28. There was no check for failure on the call to getsockname() in the daemon
5344 code. This can fail if there is a shortage of resources on the system, with
5345 ENOMEM, for example. A temporary error is now given on failure.
5347 29. Contrary to the C standard, it seems that in some environments, the
5348 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5349 program. Exim now does this explicitly; it affects the formatting of
5350 timestamps using strftime().
5352 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5353 complaints. I've now initialized all the variables, to avoid this.
5355 32. Header lines added by a system filter were not being "seen" during
5356 transport-time rewrites.
5358 33. The info_callback() function passed to OpenSSL is set up with type void
5359 (*)(SSL *, int, int), as described somewhere. However, when calling the
5360 function (actually a macro) that sets it up, the type void(*)() is
5361 expected. I've put in a cast to prevent warnings from picky compilers.
5363 34. If a DNS black list lookup found a CNAME record, but there were no A
5364 records associated with the domain it pointed at, Exim crashed.
5366 35. If a DNS black list lookup returned more than one A record, Exim ignored
5367 all but the first. It now scans all returned addresses if a particular IP
5368 value is being sought. In this situation, the contents of the
5369 $dnslist_value variable are a list of all the addresses, separated by a
5372 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5373 to accept a host name and all its aliases if the forward lookup for any of
5374 them yielded the IP address of the incoming connection. Now it accepts only
5375 those names whose forward lookup yields the correct IP address. Any other
5376 names are discarded. This closes a loophole whereby a rogue DNS
5377 administrator could create reverse DNS records to break through a
5378 wildcarded host restriction in an ACL.
5380 37. If a user filter or a system filter that ran in a subprocess used any of
5381 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5382 the wrong values were passed to the pipe command ($thisaddress had the
5383 value of $0, $0 had the value of $1, etc). This bug was introduced by
5384 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5386 38. Improved the line breaking for long SMTP error messages from ACLs.
5387 Previously, if there was no break point between 40 and 75 characters, Exim
5388 left the rest of the message alone. Two changes have been made: (a) I've
5389 reduced the minimum length to 35 characters; (b) if it can't find a break
5390 point between 35 and 75 characters, it looks ahead and uses the first one
5391 that it finds. This may give the occasional overlong line, but at least the
5392 remaining text gets split now.
5394 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5395 file descriptors might be low, and that setting 1000 would always raise it.
5396 It turns out that in some environments, the limit is already over 1000 and
5397 that lowering it causes trouble. So now Exim takes care not to decrease it.
5399 40. When delivering a message, the value of $return_path is set to $sender_
5400 address at the start of routing (routers may change the value). By an
5401 oversight, this default was not being set up when an address was tested by
5402 -bt or -bv, which affected the outcome if any router or filter referred to
5405 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5406 log only when "message" or "log_message" are set. However, if one of the
5407 conditions was an address verification, or a call to a nested ACL, the
5408 messages generated by the underlying test were being passed through. This
5409 no longer happens. The underlying message is available in $acl_verify_
5410 message for both "message" and "log_message" expansions, so it can be
5411 passed through if needed.
5413 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5414 new expansion $bh_ to give the encoded byte string without charset
5415 translation. Translation happens only if iconv() is available; HAVE_ICONV
5416 indicates this at build time. HEADERS_CHARSET gives the charset to
5417 translate to; headers_charset can change it in the configuration, and
5418 "headers charset" can change it in an individual filter file.
5420 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5421 that creates RFC 2047 encoded "words" labels them as that charset instead
5422 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5423 expansion operator; (ii) when Exim creates a From: line for a local
5424 message; (iii) when a header line is rewritten to include a "phrase" part.
5426 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5427 buggy, causing it to skip the first lines of messages whose message ID
5428 ended in 'D'. This would not have bitten before Exim release 4.14, because
5429 message IDs were unlikely to end in 'D' before then. The effect was to have
5430 incorrect size information for certain domains.
5432 45. #include "config.h" was missing at the start of the crypt16.c module. This
5433 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5436 46. If there was a timeout during a "random" callout check, Exim treated it as
5437 a failure of the random address, and carried on sending RSET and the real
5438 address. If the delay was just some slowness somewhere, the response to the
5439 original RCPT would be taken as a response to RSET and so on, causing
5440 mayhem of various kinds.
5442 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5443 when I implemented it. It didn't allow for the fact that some option values
5444 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5445 the right test for positive values.
5447 48. Domain names in DNS records are case-independent. Exim always looks them up
5448 in lower case. Some resolvers return domain names in exactly the case they
5449 appear in the zone file, that is, they may contain uppercase letters. Not
5450 all resolvers do this - some return always lower case. Exim was treating a
5451 change of case by a resolver as a change of domain, similar to a widening
5452 of a domain abbreviation. This triggered its re-routing code and so it was
5453 trying to route what was effectively the same domain again. This normally
5454 caused routing to fail (because the router wouldn't handle the domain
5455 twice). Now Exim checks for this case specially, and just changes the
5456 casing of the domain that it ultimately uses when it transmits the message
5459 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5462 50. If a filter generated a file delivery with a non-absolute name (possible if
5463 no home directory exists for the router), the forbid_file option was not
5466 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5467 the existing equality matching.
5469 52. Exim was using ints instead of ino_t variables in some places where it was
5470 dealing with inode numbers.
5472 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5473 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5474 called TMPDIR, and if it finds it is different, it changes its value.
5476 54. The smtp_printf() function is now made available to local_scan() so
5477 additional output lines can be written before returning. There is also an
5478 smtp_fflush() function to enable the detection of a dropped connection.
5479 The variables smtp_input and smtp_batched_input are exported to
5482 55. Changed the default runtime configuration: the message "Unknown user"
5483 has been removed from the ACL, and instead placed on the localuser router,
5484 using the cannot_route_message feature. This means that any verification
5485 failures that generate their own messages won't get overridden. Similarly,
5486 the "Unrouteable address" message that was in the ACL for unverifiable
5487 relay addresses has also been removed.
5489 56. Added hosts_avoid_esmtp to the smtp transport.
5491 57. The exicyclog script was not checking for the esoteric option
5492 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5493 will work only if exicyclog is run under the appropriate euid.
5495 58. Following a discussion on the list, the rules by which Exim recognises line
5496 endings on incoming messages have been changed. The -dropcr and drop_cr
5497 options are now no-ops, retained only for backwards compatibility. The
5498 following line terminators are recognized: LF CRLF CR. However, special
5499 processing applies to CR:
5501 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5502 nor a local message in the state where . is a terminator.
5504 (ii) If a bare CR is encountered in a header line, an extra space is added
5505 after the line terminator so as not to end the header. The reasoning
5506 behind this is that bare CRs in header lines are most likely either
5507 to be mistakes, or people trying to play silly games.
5509 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5510 was being incorrectly given as 18 bytes larger than it should have been.
5511 This is a VOB (very old bug).
5513 60. This may never have affected anything current, but just in case it has:
5514 When the local host is found other than at the start of a list of hosts,
5515 the local host, those with the same MX, and any that follow, are discarded.
5516 When the list in question was part of a longer list of hosts, the following
5517 hosts (not currently being processed) were also being discarded. This no
5518 longer happens. I'm not sure if this situation could ever has previously
5521 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5524 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5525 header. This is something that is recommended in a new Internet Draft, and
5526 is something that is documented as being done by Sendmail. There are two
5527 possible values. For messages generated by the autoreply transport, Exim
5530 Auto-Submitted: auto-replied
5532 whereas for all other generated messages (e.g. bounces) it adds
5534 Auto-Submitted: auto-generated
5536 63. The "personal" condition in filters now includes a test for the
5537 Auto-Submitted: header. If it contains the string "auto-" the message it
5538 not considered personal.
5540 64. Added rcpt_include_affixes as a generic transport option.
5542 65. Added queue_only_override (default true).
5544 66. Added the syslog_duplication option.
5546 67. If what should have been the first header line of a message consisted of
5547 a space followed by a colon, Exim was mis-interpreting it as a header line.
5548 It isn't of course - it is syntactically invalid and should therefore be
5549 treated as the start of the message body. The misbehaviour could have
5550 caused a number of strange effects, including loss of data in subsequent
5551 header lines, and spool format errors.
5553 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5554 client host had authenticated. This control can now be exercised by an ACL
5555 for more flexibility.
5557 69. By default, callouts do not happen when testing with -bh. There is now a
5558 variant, -bhc, which does actually run the callout code, including
5559 consulting and updating the callout cache.
5561 70. Added support for saslauthd authentication, courtesy of Alexander
5564 71. If statvfs() failed on the spool or log directories while checking their
5565 size for availability, Exim confusingly gave the error "space shortage".
5566 Furthermore, in debugging mode it crashed with a floating point exception.
5567 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5568 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5569 serious problem, Exim now writes to the main and panic logs when this
5570 happens, with details of the failure. It then refuses to accept the
5571 incoming message, giving the message "spool directory problem" or "log
5572 directory problem" with a 421 code for SMTP messages.
5574 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5575 0, 1, and 2 exist, because some OS complain for execs without them (see
5576 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5577 descriptors. However, the code omitted to check that the open succeeded,
5578 causing mysterious errors if for some reason the permissions on /dev/null
5579 got screwed. Now Exim writes a message to the main and panic logs, and
5580 bombs out if it can't open /dev/null.
5582 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5583 interact so that it is all more flexible. It is supposed to remain
5584 backwards compatible. Also added extra_local_interfaces.
5586 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5587 to bomb out with an assertion failure - to the client this appears as a
5588 connection drop. This problem occurs in the part of the code that was taken
5589 from the Samba project. Fortunately, the assertion is in a very simple
5590 function, so I have fixed this by reproducing the function inline in the
5591 one place where it is called, and arranging for authentication to fail
5592 instead of killing the process with assert().
5594 75. The SPA client code was not working when the server requested OEM rather
5595 than Unicode encoding.
5597 76. Added code to make require_files with a specific uid setting more usable in
5598 the case where statting the file as root fails - usually a non-root-mounted
5599 NFS file system. When this happens and the failure is EACCES, Exim now
5600 forks a subprocess and does the per-uid checking as the relevant uid.
5602 77. Added process_log_path.
5604 78. If log_file_path was not explicitly set, a setting of check_log_space or
5605 check_log_inodes was ignored.
5607 79. If a space check for the spool or log partitions fails, the incident is now
5608 logged. Of course, in the latter case the data may get lost...
5610 80. Added the %p formatting code to string_format() so that it can be used to
5611 print addresses in debug_print(). Adjusted all the address printing in the
5612 debugging in store.c to use %p rather than %d.
5614 81. There was a concern that a line of code in smtp_in.c could overflow a
5615 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5616 initially expressed, the concern was not well-founded, because trailing
5617 spaces are removed early. However, if the trailing spaces were followed by
5618 a NULL, they did not get removed, so the overflow was possible. Two fixes
5621 (a) I re-wrote the offending code in a cleaner fashion.
5622 (b) If an incoming SMTP command contains a NULL character, it is rejected
5625 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5626 runs initgroups(), so that if the Exim user is in any additional groups,
5627 they will be used during message reception.
5633 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob